Upload
dokhanh
View
220
Download
0
Embed Size (px)
Citation preview
Surface Water Quality Assessment and Modelling A case study in the Tuul River, Ulaanbaatar city,
Mongolia
Ochir Altansukh
March 2008
Surface Water Quality Assessment and Modelling A case study in the Tuul River, Ulaanbaatar city,
Mongolia
By
Ochir Altansukh
Thesis submitted to the International Institute for Geo-information Science and Earth Observation in
partial fulfilment of the requirements for the degree of Master of Science in Geo-information Science
and Earth Observation, Specialisation: Integrated Watershed Modelling and Management
Thesis Assessment Board Professor. Dr. Z. Bob Su Chairman WREM Department, ITC
Associate Professor. Dr. Ir. C.M.M. Chris Mannaerts First Supervisor WREM Department, ITC
Assistant Professor. Dr. Ir. Mhd. Suhyb Salama Supervisor WREM Department, ITC
Assistant Professor. Dr. Ir. D.C.M. Denie Augustijn External Examiner WEM Department, UT
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION
ENSCHEDE, THE NETHERLANDS
Disclaimer This document describes work undertaken as part of a programme of study at the International Institute for Geo-information Science and Earth Observation. All views and opinions expressed therein remain the sole responsibility of the author, and do not necessarily represent those of the institute.
To my father, Ochir’s family.
You are all special to me.
i
Acknowledgements
First of all, my deepest thanks to my father, Ochir’s family, including my mother, my sisters, my
brothers, my wife and my son, my younger sisters, my younger brothers for their honourable support,
encouragements and love.
Specially, all my infinite love, sincere gratitude and unconditional faith to my Mandaa for her pure
heart, unlimited support and eternal love.
Special thanks to Dr. J.L. John van Genderen, who is a professor at ITC, Department of Earth
Observation Science, for his vital support to “open the gate” and kind-heart to the Geo-science in
Mongolia.
I wish to express my honest gratitude to Ms. D.Narantuya, a leader of NGIC for NRM project, and to
Mr. D.Amarsaihan, a training coordinator of NRM project, for their support to my research and
extension of study.
I am grateful to Ir. A.M. Arno van Lieshout, a programme director of WREM, and Dr. H.A.M.J. Hein
van Gils, a professor at ITC, NRM, who gave me very nice opportunity to study at ITC in MSc
course.
Datasets used in this study have been obtained from different organizations and persons, including
NAMHEM, CLEM, Ms. Munhtsetseg (Tuul-UB hydrological station), Mr. G.Davaa (WS,
NAMHEM), Ms. Ya.Erdenebayar (CLEM) and Mr. J.Bayasgalan (NGIC project). We would like to
thank them all for making available data.
I do particular thank to professor Dr. Ir. C.M.M. Chris Mannaerts, my supervisor, for his time,
comments, valuable ideas, support and guidance throughout the research. I wish to thank Dr. Ir. Mhd.
Suhyb Salama, my supervisor, for his ideal comments and time to read my thesis.
I appreciate to all my country mates, Ms. B.Orgil, Ms. B.Oyundari, Ms. M.Munguntuya, Mr.
D.Bayarbaatar, Ms. B.Tuul, and my classmates and friends, above all Ms. Le Thi Hanh, Ms. Nguyen
Phuong Tam. I will never forget your support, friendship and very nice time that we were shared
together at ITC.
Thanks to all ITC staffs, especially to WREM staffs for their important support and very advanced
lectures, classes during my study.
On top of all, thank very much to my God.
iii
Abstract
This research was conducted to assess the stream water quality of the Tuul River around the
Ulaanbaatar city, Mongolia and to model the impacts of wastewater treatment plants on the Tuul
River water quality, hereby mainly focusing on the CWTP (Central Wastewater Treatment Plant).
The assessment of surface water quality for natural river waters, was done using a water quality index,
developed by Erdenebayar,Ya., and Bulgan,T and the Mongolian National Standard MNS 4586-98.
This national standard was enacted by the CSM, Centre of Standardization and Measurement in 1998,
and represents the standards for maximum permissible levels of chemical variables in the surface
waters in Mongolia.
Geo-statistical techniques were utilized to estimate the spatial and temporal variability of the surface
water quality index, which was calculated by combination of 6 parameters, i.e. ammonia, nitrate,
nitrite, dissolved oxygen, chemical oxygen demand and biochemical oxygen demand. The monitoring
period covered 11 years from 1996 to 2006. The time series of water quality maps for the Tuul River
were also visualized using ILWIS.
DMS (Duflow Modelling System) was then used as tool for modelling of the river water quality. The
DO (dissolved oxygen) in the Tuul River network system was selected as main state variable. The DO
was modelled, calibrated and validated using an oxygen quality model (WUR, 2002) permitting to
analyse oxygen depletion by nitrogen and carbonaceous waste inputs in the river. Water quality
simulations were performed using hydro-chemical, hydraulic, climatic datasets between 2005 and
2006. The calibrated water quality model was then applied to evaluate the impact scenarios (of the
CWTP of Ulaanbaatar), for improvement of river water quality in order to meet the Mongolian
National Standards.
Keywords;
river water quality assessment, chemical datasets, hydraulic data, climatic data, pollution map,
contamination source, flow model, quality model, model calibration, model validation, sensitivity
analysis
v
Table of contents
Acknowledgements .................................................................................................................................. i Abstract .................................................................................................................................................. iii List of tables.......................................................................................................................................... vii List of figures....................................................................................................................................... viii List of equations.......................................................................................................................................x List of appendices .................................................................................................................................. xi Abbreviations ........................................................................................................................................ xii Chapter 1: Introduction ............................................................................................................................1
1.1: Background information ...............................................................................................................1 1.1.1: Spatial information ................................................................................................................1 1.1.2: Hydraulic information ...........................................................................................................2 1.1.3: Climatic information..............................................................................................................4 1.1.4: Hydro-chemical information..................................................................................................5
1.2: Recent problem statement.............................................................................................................5 1.3: Reason of topic selection ..............................................................................................................6 1.4: Research object .............................................................................................................................7 1.5: Research targets ............................................................................................................................7 1.6: Previuos studies ............................................................................................................................7
1.6.1: Quality studies using chemical method .................................................................................8 1.6.2: Quality studies using biological method................................................................................8
1.7: Research objectives.......................................................................................................................9 1.8: Research questions......................................................................................................................10 1.9: Research phases ..........................................................................................................................10 1.10: Methodology .............................................................................................................................11 1.11: Research hypothesis..................................................................................................................12 1.12: Outline of the thesis ..................................................................................................................12
Chapter 2: Field survey and data collection...........................................................................................13 2.1: Field survey.................................................................................................................................13
2.1.1: Selection of sampling site....................................................................................................13 2.1.2: Sampling method .................................................................................................................15 2.1.3: Chemical analysis in field....................................................................................................15
2.2: Data collection ............................................................................................................................16 2.2.1: Chemical data ......................................................................................................................16 2.2.2: Hydraulic data......................................................................................................................17 2.2.3: Climatic data........................................................................................................................17 2.2.4: Map and standards ...............................................................................................................18 2.2.5: Pollution source ...................................................................................................................19
Chapter 3: Water quality assessment .....................................................................................................21 3.1: Method for surface water quality assessment .............................................................................21 3.2: Basic geo-statistical analysis ......................................................................................................22
3.2.1: Descriptive statistics ............................................................................................................22 3.2.2: Exploratory graphics............................................................................................................24
3.3: Spatial water quality assessment.................................................................................................25 3.4: Temporal water quality assessment ............................................................................................27
3.4.1: Seasonal water quality assessment ......................................................................................30 3.5: Chapter conclusions....................................................................................................................31 3.6: Limitation in WQA.....................................................................................................................32
vi
Chapter 4: River flow model ................................................................................................................. 33 4.1: General introduction of flow model........................................................................................... 33 4.2: Background theory of flow simulation....................................................................................... 33
4.2.1: Saint-Venant equation......................................................................................................... 33 4.2.2: Manning equation ............................................................................................................... 34 4.2.3: Froude number .................................................................................................................... 35 4.2.4: Nash-Sutcliffe coefficient ................................................................................................... 36 4.2.5: Statistical methods ME, MAE, RMSE ............................................................................... 37
4.3: Flow model setup ....................................................................................................................... 37 4.3.1: Network development ......................................................................................................... 38 4.3.2: Initial condition of flow model ........................................................................................... 40 4.3.3: Boundary condition of flow model ..................................................................................... 41 4.3.4: Calculation setting .............................................................................................................. 42
4.4: Calibration.................................................................................................................................. 42 4.4.1: Sensitivity analysis.............................................................................................................. 44
4.5: Validation................................................................................................................................... 45 4.6: Chapter conclusion..................................................................................................................... 46 4.7: Limitation in a flow modelling .................................................................................................. 46
Chapter 5: Water quality model ............................................................................................................ 47 5.1: Introduction of quality model..................................................................................................... 47 5.2: Background theory of quality modelling ................................................................................... 48
5.2.1: Mass transport equation ...................................................................................................... 48 5.2.2: Peclet number...................................................................................................................... 50 5.2.3: DO balance equations ......................................................................................................... 50
5.3: Quality model setup ................................................................................................................... 53 5.3.1: Quality description.............................................................................................................. 55 5.3.2: Initial condition of quality model ....................................................................................... 55 5.3.3: Boundary condition of quality model ................................................................................. 56 5.3.4: Parameters in quality model................................................................................................ 57 5.3.5: External variables................................................................................................................ 57 5.3.6: Calculation setting .............................................................................................................. 58
5.4: Calibration.................................................................................................................................. 58 5.4.1: Sensitivity analysis.............................................................................................................. 59
5.4.1.1: Very sensitive parameters ............................................................................................ 59 5.4.1.2: Less sensitive parameters............................................................................................. 60 5.4.1.3: Very less and non sensitive parameters ....................................................................... 61
5.5: Validation................................................................................................................................... 62 5.5.1: Scenarios for SWQ improvement ....................................................................................... 63
5.6: Chapter conclusion..................................................................................................................... 64 5.7: Limitation in quality modelling ................................................................................................. 64
Chapter 6: Conclusions and recommendations ..................................................................................... 65 6.1: Overall conclusions.................................................................................................................... 65 6.2: Recommendations ...................................................................................................................... 66 6.3: Future research........................................................................................................................... 67 6.4: Output significance .................................................................................................................... 67 6.5: Limitations of the study ............................................................................................................. 67
References ............................................................................................................................................. 69 Appendices ............................................................................................................................................ 73
vii
List of tables
Table 1: Annual mean water balance of the Tuul River at Ulaanbaatar station ......................................2 Table 2: Area of sub-catchments of the Tuul River.................................................................................3 Table 3: Percentage of runoff components ..............................................................................................3 Table 4: Annual average discharges of runoff components in m3 s-1.......................................................3 Table 5: Long-term mean values of meteorological variables (1965 - 2000)..........................................4 Table 6: Monthly average lumped rainfall in the Tuul River basin.........................................................4 Table 7: Data, instrument requirement and sources...............................................................................11 Table 8: Spatial and temporal information of water quality sampling points .......................................14 Table 9: Data collection and sources .....................................................................................................16 Table 10: Hourly average solar radiation in W h/m2 .............................................................................18 Table 11: Permissible level of surface water variables..........................................................................19 Table 12: Assessment of surface water quality......................................................................................21 Table 13: Definition for classification of surface water quality in Mongolia .......................................22 Table 14: Statistical summary of water quality index ...........................................................................23 Table 15: The quantity of samples with critical values .........................................................................23 Table 16: Annual mean water quality index ..........................................................................................31 Table 17: Initial condition in flow model ..............................................................................................41 Table 18: Statistical evaluation of sensitivity analysis ..........................................................................44 Table 19: Initial condition in quality model ..........................................................................................56 Table 20: Quality boundary condition ...................................................................................................56 Table 21: Statistical evaluation of very sensitive parameters................................................................60 Table 22: Statistical evaluation of less sensitive parameters.................................................................61 Table 23: Statistical evaluation of not sensitive parameters..................................................................61 Table 24: Scenario for SWQ improvement, Option 1 ...........................................................................63 Table 25: Scenario for SWQ improvement, Option 2 ...........................................................................64 Table 26: Scenario for SWQ improvement, Option 3 ...........................................................................64
viii
List of figures
Figure 1: Elevation map of the Ulaanbaatar city..................................................................................... 1 Figure 2: Climadiagram at the Ulaanbaatar station................................................................................. 5 Figure 3: Location of research field ........................................................................................................ 7 Figure 4: Methodology flowchart ......................................................................................................... 12 Figure 5: Photo at Terelj-Terelj sampling point.................................................................................... 13 Figure 6: Locations of sampling points................................................................................................. 14 Figure 7: Monthly mean water temperature.......................................................................................... 17 Figure 8: Monthly average wind speed at Ulaanbaatar station ............................................................. 17 Figure 9: Data frame of the Tuul River dataset..................................................................................... 23 Figure 10: Highest value of WQI.......................................................................................................... 23 Figure 11: Histogram and box plot of water quality index ................................................................... 24 Figure 12: Box plot for spatial variability of WQI................................................................................ 24 Figure 13: Correlation between distance & WQI.................................................................................. 25 Figure 14: Correlation between distance and WQI in the upstream section......................................... 26 Figure 15: Water quality fluctuation along the upstream in 1996-2006 ............................................... 26 Figure 16: Correlation between distance and WQI in the downstream section.................................... 27 Figure 17: Correlation between time & WQI (1996-2006)................................................................... 27 Figure 18: Water quality fluctuation at Tuul-Uubulan SP during study period.................................... 28 Figure 19: Water quality fluctuation at Tuul-Songino (upper) SP during study period ....................... 28 Figure 20: Water quality fluctuation at Tuul-Songino (down) SP during study time........................... 28 Figure 21: Water quality along the Tuul River in 1996 ........................................................................ 29 Figure 22: Water quality along the Tuul River in 2002 ........................................................................ 29 Figure 23: Water quality along the Tuul River in 2006 ........................................................................ 30 Figure 24: Seasonal water quality in upstream reach............................................................................ 30 Figure 25: Seasonal water quality in downstream reach....................................................................... 30 Figure 26: Correlation between river monthly mean discharge and monthly WQI.............................. 32 Figure 27: Flowchart of a flow model................................................................................................... 38 Figure 28: A flow model network ......................................................................................................... 39 Figure 29: Simplified cross-section of the Tuul River under Zaisan Bridge ........................................ 39 Figure 30: Simplified cross-section of the Tuul River under Altanbulag Bridge ................................. 40 Figure 31: Discharge of the Tuul River under Zaisan Bridge in 2005 and 2006.................................. 41 Figure 32: Artificial discharge at end node........................................................................................... 42 Figure 33: Correlation between calculated and observed discharge in calibration year....................... 43 Figure 34: Observed and modelled Q in calibration year ..................................................................... 43 Figure 35: Sensitivity analysis .............................................................................................................. 44 Figure 36: Correlation between calculated and observed Q data in validation year............................. 45 Figure 37: Observed and modelled Q in validation year....................................................................... 45 Figure 38: Schematization of DO balance model ................................................................................. 50 Figure 39: Flowchart of a quality model ............................................................................................... 54 Figure 40: Quality model network schematization ............................................................................... 54 Figure 41: Correlation between calculated and observed DO in calibration year ................................ 58
ix
Figure 42: Observed and modelled DO in calibration year ...................................................................59 Figure 43: Sensitivity analysis for beta..................................................................................................59 Figure 44: Sensitivity analysis for SOD ................................................................................................60 Figure 45: Sensitivity analysis for Vs....................................................................................................61 Figure 46: Correlation between calculated and observed Q data in validation year .............................62 Figure 47: Observed and modelled Q in validation year .......................................................................62 Figure 48: Options for SWQ improvement............................................................................................63
x
List of equations
Equation 1: Surface water quality index ............................................................................................... 21 Equation 2: Mass equation .................................................................................................................... 33 Equation 3: Momentum equation.......................................................................................................... 33 Equation 4: Discharge equation ............................................................................................................ 34 Equation 5: Correction factor for non-uniformity of velocity distribution........................................... 34 Equation 6: Hydraulic radius................................................................................................................. 34 Equation 7: Slope of riverbed................................................................................................................ 35 Equation 8: River velocity..................................................................................................................... 35 Equation 9: Froude number................................................................................................................... 35 Equation 10: Hydraulic depth ............................................................................................................... 35 Equation 11: Nash-Sutcliffe coefficient................................................................................................ 36 Equation 12: Mean error ....................................................................................................................... 37 Equation 13: Mean absolute error ......................................................................................................... 37 Equation 14: Root means squared error ................................................................................................ 37 Equation 15: Slope angle equation........................................................................................................ 40 Equation 16: Momentum balance equation........................................................................................... 48 Equation 17: Mass balance equation..................................................................................................... 49 Equation 18: Equation for the constituent transport by advection and dispersion ............................... 49 Equation 19: Background dispersion coefficient .................................................................................. 49 Equation 20: Shear stress equation ....................................................................................................... 49 Equation 21: Peclet number .................................................................................................................. 50 Equation 22: Dissolved oxygen balance ............................................................................................... 50 Equation 23: Temperature dependent oxygen mass-transfer velocity .................................................. 51 Equation 24: Re-aeration rate................................................................................................................ 51 Equation 25: Oxygen saturation............................................................................................................ 51 Equation 26: Primary production .......................................................................................................... 52 Equation 27: Oxygen consumption for sediment .................................................................................. 52 Equation 28: Ultimate BOD.................................................................................................................. 52 Equation 29: Biochemical oxygen demand........................................................................................... 53 Equation 30: Oxygen consumption for nitrification ............................................................................. 53 Equation 31: Dispersion coefficient...................................................................................................... 57
xi
List of appendices
Appendix 1: Photos in the field .............................................................................................................73 Appendix 2: Note for water field survey ...............................................................................................75 Appendix 3: Field measurements...........................................................................................................76 Appendix 4: Hydro-chemical dataset.....................................................................................................78 Appendix 5: Hydraulic dataset...............................................................................................................80 Appendix 6: CWTP chemical dataset ....................................................................................................82 Appendix 7: A flow model setup ...........................................................................................................83 Appendix 8: Manning value...................................................................................................................84 Appendix 9: Quality model syntax ........................................................................................................85 Appendix 10: A quality model setup .....................................................................................................86 Appendix 11: Required input data for the quality model ......................................................................87
xii
Abbreviations
ALAGC Administration of Land Affairs, Geodesy and Cartography
BOD5 Biochemical Oxygen Demand (5 day)
COD-Mn Chemical Oxygen Demand (Manganese III method)
CSM Centre of Standardization and Measurement
CWTP Central Wastewater Treatment Plant
DMS Duflow Modelling System
DO Dissolved Oxygen
ITC International Institute for Geo-Information Science and Earth Observation
IWEC International Weather for Energy Calculations
JICA Japanese International Cooperation Agency
MAE Mean absolute error
MASM Mongolian Administration of Standardization and Measurement
ME Mean error
MNE Ministry of Nature and Environment
NAMHEM National Agency of Meteorology, Hydrology and Environment Monitoring
NGIC for NRM National Geo-Information Centre for Natural Resource Management
NGO Non-governmental Organization
NSA National Standard Agency
NUM National University of Mongolia
RMSE Root mean squared error
SCLM State Central Library of Mongolia
SP Sampling point
SWQ Surface Water Quality
SWQI Surface Water Quality Indices
UB Ulaanbaatar city
WQA Water Quality Assessment
WQI Water Quality Index
WSSA Water supply and sewerage authority
WTS Wastewater Treatment Station
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 1
Chapter 1: Introduction
1.1: Background information
The background information of study area consists of four different parts.
1.1.1: Spatial information
Mongolia is located in the Central Asia between 41035’ - 52006’ North and 87047’ - 119057’ East.
Mongolian overall territory is 1.57 million square kilometres, total width is 1260 km from north to
south and total length is 2370 km from west to east.
Mongolia is situated in arid and semi-arid natural zone. The climate is harsh, with less
precipitation (approximately 200 mm year-1) and greatly fluctuating temperatures varying between
-350C in January and 320C in July.
The main type of land usage is, because of less rainfall, rangeland for nomadic livestock and
husbandry, fodder crop productions are of minor importance, the limiting factor is the lack of
water for irrigation. The total surface water resources in Mongolia is estimated 599 km3 year-1 and
is composed of water stored in lakes 500 km3 year-1, glaciers 62,9 km3 year-1 and rivers
34.6 km3 year-1 [Baasandorj and Davaa, 2006].
Ulaanbaatar, capital city of Mongolia, is placed in between E 106043’ - E 107002’ of longitude and
between N 47053’ - N 47057’ of latitude. Elevation ranges between 1214 m to 2079 m above mean
sea level and stretches from SW to NE.
Figure 1: Elevation map of the Ulaanbaatar city
Source: SRTM [SRTM, 2005]
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 2
In territory of the Ulaanbaatar city, has about 49 streams and rivers (most of them are dried up).
Three rivers of them, which are named Selbe, Uliastai and Tuul, are running through the central
part of the Ulaanbaatar city. Two rivers are tributaries of the Tuul River, which is the biggest one.
The origins of the Tuul River are the Namiya and the Nergui streams at the southwestern slope of
the Khentii mountain range that is located in approximately 2000 meters above the mean sea level.
Terrain of the Tuul River basin ranges from 1200-2700 meters above mean sea level [Dashdeleg
and Bat, 1971].
The total length of the Tuul River is around 720 km and flows through the Ulaanbaatar city. Its
catchment area covers roughly 49,840 km2 (3.19% of the entire territory of Mongolia) and consists
of 11 soums of five aimags1 territories, which are namely Arhangai, Uvurhangai, Selenge, Bulgan
and Tuv. The Tuul River, located in the heart of the Ulaanbaatar city, is an environmentally,
economically and socially significant natural resource [Roza-Butler, 2004]. Almost hundreds of
tourist camps are to be found around the runoff formation zone and the gold mining activities are
taking place along the downstream of the Tuul River [Orgil, 2007].
1.1.2: Hydraulic information
Totally, eight hydrological stations were sited along the river basin. Four of them are stopped and
two of them are half-operational due to old technique and economic constraint. Rest two stations
are functioning, nowadays. The oldest hydrological station is “Tuul-Ulaanbaatar” which is
operating since 1942.
Table 1: Annual mean water balance of the Tuul River at Ulaanbaatar station
Precipitation 250 mm
Discharge 149 mm
Evapotranspiration 101 mm
Source: NAMHEM
The Tuul River basin can be divided into following eight sub-basins:
− Upper Tuul basin
− Terelj basin
− Khol basin
− Uliastai basin
− Selbe basin
− Turgen basin
− Middle-lower Tuul
− Kharuuh basin
Biggest tributary of the Tuul River in upper basin is the Terelj River that is draining from
catchment area of 1380.4 km2. After confluence of rivers Turgen and Tuul, there is nearly no
tributary in middle-lower Tuul sub-catchment. Size of sub-catchments area have been shown in
Table 2 [NAMHEM, 1999].
1 Soum and aimag are names of the administrative units in Mongolia
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 3
Table 2: Area of sub-catchments of the Tuul River
No Name of the sub-catchments Size of the catchment area (km2)
1 Upper Tuul 2698.5
2 Terelj 1380.4
3 Khol 249.5
4 Uliastai 308.9
5 Selbe 303.4
6 Turgen 531.6
7 Kharuukh 16576.2
Runoff components of the Terelj and the Tuul rivers were separated by hydrological method.
Hydrological and meteorological observations record that spring flood starts in middle of April and
low flow period occurs in June. Annual runoff of the Tuul River consists of the following three
contributors: (i) 69% from the rainfall, (ii) 26% from groundwater and (iii) 5% from snow. The
spatial distribution of groundwater contribution decreases along the Tuul River. This is because
80% of the Ulaanbaatar city’s water supply is provided by groundwater. Meanwhile, contribution
of precipitation increases in downstream with increment of catchment area [Baasandorj and
Davaa, 2006].
Table 3: Percentage of runoff components
Stations Year Groundwater Snow water Rainfall
Tuul – Bosgo bridge 2000 44.7 5.3 50.0
Terelj - Terelj 2000 41.6 5.2 53.3
Tuul – Ulaanbaatar 2000 37.4 7.3 55.2
The average channel width of the Tuul River is 35 to 75 meters during non-flooding time, depth is
0.8-3.5 m and the velocity is 0.5-1.5 m s-1. The annual mean flow of the Tuul River is
approximately 26.6 m3 s-1. The observed maximum discharge reached to 1580 m3 s-1 and 564 m3 s-1
in Ulaanbaatar and Terelj stations, respectively. During the low flow period of warm season, it has
dropped until 1.86 m3 s-1 at Ulaanbaatar and 0.44 m3 s-1 at Terelj stations [NAMHEM, 1999].
Table 4: Annual average discharges of runoff components in m3 s-1
Stations Year Groundwater Snow water Rainfall Annual mean flow
Tuul – Bosgo bridge 2000 6.06 0.73 6.78 13.6
Terelj - Terelj 2000 2.70 0.34 6.08 9.12
Tuul – Ulaanbaatar 2000 4.90 0.96 7.23 13.1
Water demand of the city has increased by 20% from 1998 to 2005. Population growth,
urbanization and intensity of industries have created water exploitation, deterioration of natural
water regime and ecological degradation of the Tuul River basin. Moreover, the Tuul River drains
into the Orhon River, one of the main tributaries of the Selenge River. The Selenge River is the
main tributary of the Lake Baikal in Russia, the world largest freshwater lake by volume. The Tuul
River basin is economically most important and one of the mainly polluted rivers in Mongolia.
However, no management plan currently exists for the water resources of the Tuul River basin
[Orgil, 2007].
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 4
1.1.3: Climatic information
In the Tuul River basin, has continental climatic feature that is characterized by wide variations of
annual, monthly and daily temperatures, low range of air humidity, non-uniform distribution of
precipitation, cold, long-lasting winter and warm summer. The rainy period continues from June to
August in the upper Tuul River basin, of which rainfall shares about 74% of the annual
precipitation.
Annual average air temperature is -1.20C in the study area. Annual minimum temperature is
recorded -39.60C in January, maximum temperature reaches 34.50C. Fluctuation of air
temperatures reaches approximately 74.10C.
Table 5: Long-term mean values of meteorological variables (1965 - 2000)
Variables Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
Precipitation, mm 2.2 1.6 3.7 7.3 14.7 54.6 57.9 75.9 23.4 9.7 4.2 3.2 258.5
Air temperature, 0C -21.8 -18.1 -8.8 0.8 9.4 14.9 16.9 14.8 8.3 0.1 -11.4 -19.5 -1.2
Wind speed, m s-1 1.6 2.1 2.9 3.9 4.4 3.5 3.1 2.6 2.9 2.6 2.0 1.6 2.8
Annual mean precipitation is 258.5 mm at Ulaanbaatar station and almost 90 percent occurs in
warm session of year, particularly in April to September. The mean value of precipitation, which
occurs in warm period, is 233.8 mm that is falling dominantly in form of thunderstorm. Daily
maximum precipitation occurred in 1967 that recorded 74.9 mm [Baasandorj and Davaa, 2006].
Table 6: Monthly average lumped rainfall in the Tuul River basin
Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Cumulated
Buyant-Uhaa 1.48 1.40 2.3 6.29 13.72 43.29 68.00 73.73 32.66 9.15 6.82 2.85 258.5
Tahilt 1.79 1.43 2.64 7.59 13.40 48.77 69.90 76.20 32.78 8.69 4.32 3.07 270.5
MUIS 1.93 1.64 4.54 9.28 14.22 49.00 72.04 75.54 31.30 9.62 4.87 2.92 276.8
In the study area, wind direction dominants from north and northwestern. However, surrounding
mountains make change in locally. Annual mean wind speed is recorded 2.8 m s-1. The 53.6% of
annual wind speed falls in range of 0-2 m s-1, 26.8% falls 2-5 m s-1, 12.5% belongs to 6-10 m s-1
and only 1.3% concerns to the 11-15 m s-1. Annual air humidity is 62% and its range in December
and January is 70-74%, lowest value is 48% in April. Average vapour pressure is 4.8 and
saturation deficit is 3.8 hPa. Snow in the study area covers 68 days in average. It begins from
October and stabilizes in middle of November, melts in May [Baasandorj and Davaa, 2006].
Climadiagram of study area at Ulaanbaatar meteorological station has been shown by the
following graph.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 5
Figure 2: Climadiagram at the Ulaanbaatar station
-30
-20
-10
0
10
20
30
40
50
Jan-
04F
eb-0
4M
ar-0
4A
pr-0
4M
ay-0
4Ju
n-04
Jul-0
4A
ug-0
4S
ep-0
4O
ct-0
4N
ov-0
4D
ec-0
4Ja
n-05
Feb
-05
Mar
-05
Apr
-05
May
-05
Jun-
05Ju
l-05
Aug
-05
Sep
-05
Oct
-05
Nov
-05
Dec
-05
Jan-
06F
eb-0
6M
ar-0
6A
pr-0
6M
ay-0
6Ju
n-06
Jul-0
6A
ug-0
6S
ep-0
6O
ct-0
6N
ov-0
6D
ec-0
6
0C
-60
-40
-20
0
20
40
60
80
100
mm
Air temperature Precipitation
Source: Data from NAMHEM
1.1.4: Hydro-chemical information
The Tuul River quality is naturally clean and rich of calcium, bicarbonate. Total dissolved solid of
the river water ranges from 100-210 mg l-1 along its reaches. The Tuul River contains 78.9% Ca+2,
15.8% Na+, 28.1-634.7 mg l-1 mineral and it belongs to the hydro-carbonate class, calcium group,
pH = 6.1-7.5. The main cation is calcium and dominant anion is hydro-carbonate in the Tuul River.
Moreover, cation proportion is Ca+2 > Mg+2 > Na+ + K+ and anion ratio is HCO3- > SO4
-2 > CI-.
Naturally, anion and cation proportions and chemical content of water matches to the river with
pure water [NAMHEM, 1999].
Along the Tuul River, water quality has been monitored since 1980th. Chemical and biological
variables such as pH, nitrite, nitrate, ammonium, phosphate, DO, COD-Mn, and BOD5 are
analysing in the CLEM. The Tuul River is nearly not polluted until the Ulaanbaatar city and
pollution starts during running through UB. Most of tributaries, especially rural areas, are not
affected by human activities and have good self-purification rate [Baasandorj and Davaa, 2006].
Unfortunately, in the last few years, rapid urbanization and increment of industries have negatively
influenced water quality and chemical composition of river in surrounding area of the Ulaanbaatar
city [Javzan, et al., 2004]. Therefore, chemical content of river suddenly changes from west part of
the Ulaanbaatar city. The reason of chemical changes is the non-completely treated wastewater
from the Central Wastewater Treatment Plant (CWTP), which is located in west side of the
Ulaanbaatar city, pours into the Tuul River [Altansukh, 2000].
1.2: Recent problem statement
Water quality and pollution of the Tuul River has been monitored in 10 stationary points along the
river since 1980s. Fast urbanization combinations with increasing number of tourist camps,
agricultural and mining activities have had a significant negative impact on the Tuul River’s quality
and its associated ecosystems. Consequently, water becomes seriously polluted, loses its clarity,
transparency and self-purification rate decreases in year by year.
According to results of the hydrological survey was conducted in 2003, hydrological regime and its
runoff formation zones of the Tuul River are gradually being changed and polluted by the settlements,
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 6
intensive overgrazing, timbering, wild fires and improper wastewater treatment in the river banks
[Baasandorj and Davaa, 2006].
Water contamination of the Tuul River is continuously increasing from upper to its lower reach.
Naturally, upstream of the Tuul river is running through mountainously area and there has high
velocity and turbulent. Hence, upper part of river has more oxidization potential, re-aeration and self-
purification. When it comes to the Ulaanbaatar city, natural condition changes from mountainous
region to valley. In valley, velocity and turbulence of river decreases, then capability of oxidization,
re-aeration and self-purification also reduces as well. This is the natural factor of possibility to store
contaminant elements in the river water a longer time and distance [Altansukh, 2000].
In last decade, contaminant of the Tuul River is constantly increasing related to increment of
industries, agricultures and old, insufficient sewage and treatment system. Along the Tuul River has
five contamination point sources. The self-purification coefficient of the Tuul River is 6.57 until it
reaches first contamination point source and it reduces until 0.98 after the CWTP wastewater is
pouring into river. The main and biggest artificial point source of pollution in the Tuul River basin is
improper treated wastewater from the CWTP (190,000 m3 day-1), then Nalaih (1,400 m3 day-1), Niseh
(400 m3 day-1), Bio-industry (490 m3 day-1) and Bio-Songino wastewater treatment station (600 m3
day-1). The CWTP is responsible for the collection and treatment of industrial and domestic sewage
water in Ulaanbaatar. The treatment efficiency of the CWTP as well as other wastewater treatment
stations in the region is often inadequate due to financial constraint. The efficiency of the CWTP is
approximately 60-70% due to poor maintenance, lack of spare parts, outdated equipment and frequent
power shortages. This also causes of 10-20% wastewater on a daily basis to be released directly into
the Tuul River without any treatment [Orchlon, 1995]. Air, soil pollution and accumulated wastes in
catchment area, which are transferred by surface runoff and flood channel, are also have significant
impact on river water quality. Population of the Ulaanbaatar city is produced around 800-1000 tons of
dry wastes per day. Efficiency of the CWTP was 71% in 2002. This percentage has dropped to 66%
in 2003 and treatment level was lower than 50%, even sometimes was not operated in May 2003 and
April 2004. The major causes of water pollutant are mining industries in lower basin of the Tuul
River. Approximately, 179 licensed mining companies are operating in 145 km2 area of the basin
[MNE, 2006].
Stationary hydro-biological monitoring of invertebrate species along the Tuul River has started since
1997. Aquatic and native communities have disappeared due to pollution and gold mining activities
[MNE, 2006].
1.3: Reason of topic selection
Based on the following motivations, this topic selected for an MSc research.
− Data availability
− Attention from public and interest groups
− Current situation
− Research of the NGIC for NRM project, MNE
− Possibility of future study
The hydro-chemical and hydraulic data are availably, which measures in the Central Laboratory of
Environmental Monitoring (CLEM) in each month and National Agency of Meteorology, Hydrology
and Environment Monitoring in daily, respectively. However, there is no thematic map of surface
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 7
water contaminant. Therefore, this kind of research is still open in Mongolia and it has opportunity in
future study using different contaminant variables, in different location.
Moreover, the Tuul River is most contaminated river in Mongolia and it is running through capital
city, utilizes for manufacture and drinking purposes. That is why; the public, political groups
including NGO and scientific groups pay attention to this river.
This research closely related to the water management sub-component of the NGIC for NRM project,
MNE that is ongoing in Mongolia. In addition, Dutch government has funded this project.
Therefore, the Tuul River provides an appropriate case for this study.
Moreover, as the availability of more powerful digital computers, modelling techniques and software
has rapidly increased, the use of both physical and analogue models in hydrology has been largely
replaced by that of computer implemented mathematical models, which are usually cheaper and much
more flexible [Dingman, 2002].
1.4: Research object
Study objects are the Tuul River and its three tributaries, namely Terelj, Uliastai and Selbe Rivers,
which are running through the capital city of Mongolia.
Figure 3: Location of research field
1.5: Research targets
− River quality assessment and pollution mapping in above mentioned rivers (total 14 sampling
points) using nitrite NO2-, nitrate NO3
-, ammonium NH4+, BOD5, COD-Mn, DO (total 6
chemical variables)
− River flow and quality modelling using hydraulic variables (discharge, level), physical
parameter (cross-section of river channel, location, distance), climate data (wind, solar
radiation) and chemical variables (NH4+, BOD5, DO)
1.6: Previuos studies
Previous studies about the Tuul River can be divided into the following two different directions:
1. Quality studies using chemical method
2. Quality studies using biological method
#
##
##
#
##
#
##
#
#
#
Òóóë - Õà äàíõ ÿñàà
Òóóë - Àëòàí áó ë àã
Òóóë - Øóâ óó í ôàáðè ê
Íà ëàéõ
Ãà÷ óóðò
Ñî íãèí î
Óëààíáààòàð
Òóóë - Çà éñàí
Ñýë áý - Óëààíáààò àð
Òóóë - Ñîí ñã îë îí
Òóóë - Ñî íãè í û ä îî ä
Òóóë - Ñîíãèí û äýýä
Òýð ýëæ - Òýðýëæ
Òóóë - Óóáóë àí
Òóóë - Íà ë àé õ
Òóóë - Áàÿíç óð õ
Óëè àñòàé - Óë àà íáà àòàð
Äó íä ã îë - Óë àà íáàà òà ð
Í àð èé íè é ãîë
Áóð
ä ãîë
Íà ðèé íèé ãîë
Íà ðè éí èé ã îë Ìàíäàë
ûí ãîë
Õóéí
ãî ë Øàâà ðò ãîë
Õóéí ã
îë
Øèæ èðèéí ãî ë
Òàõ èë
ò ûí àð ãîë
Ò îëãî éòû í ãî ë
Ñýë áý ã
îë
Õàíäãà éòûí ã
î ë
Ñ ýëáý ã
îë
Áèë ã
ý õèé í ã
îë
Ñ ýëõ ã îë
Áèëãý õè
é í ãî ë
Ñ ýëáý ãîë
Äóíä ãî ë
Áàðó óí ñà ëàà ã îë
Çóó
í ñ àëàà
ãîë
Ò îëãîéò
ûí ãîë
Äóíä
ã î ë
Òàõèë
òûí ã
îëÁàÿí ãîë
Áàð óó í ãî ë
Óëèà ñ
ò àé í ãîë
Ç óóí ã
îë
Õîéä áà
ÿí ãîë
Ó ðä áàÿí ã îë
Óëèà ñ
ò àé í ã
îë
Óëè àñ
òàéí
ãîë
Õà÷
óóð ãîë
Áó
ãàòû í ãî ë
Øàâàð
òûí ã
îë
Øèëæèð ãîë
Ãà÷óó
ðòûí
ãîë
Èõ õó
àí äûí ã
îë
Äýí ä è
éí ãîë
Òý ðý ëæè
é í ã îë
Òýðý
ëæèéí ãîë
Òý ðý ëæ èéí ãîë
Ò ýðý ëæè éí ãîë
À ð ãîð
õè
Òóóë ãîë
Òóóë ã
îë
Ò ó ó
ë ãîë
Ýëñòè éí ãîë
Ýëñò
ãîë
Ýëñò ãî ë
Îâ îð ãîðõ è
Íàëà
éõûí
ãîë
Ò
ó óë ã îë
Õîë è
éí ãî
ë
Òî ð õóð
õû í ãîë
Á óìá àò
û í ã îð õè
Õîëè é í ãîë
Ò óó ë ãî ë
Ò ó ðã ýí èé ã
îë
Îâîð ç à
é ñà íãèéí ã
îë
Òóðãý íèé
ãîë
Áóõýã ãîë
Òóó ë ãîë
Ò óóë ãîë
Òóóë ãîë
Òóóë ãîë
Territory of th eTehnogen are aWater pollutio nWater pollutio nWater pollutio nStudying river .Unstudying ri v
# Research poi n
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 8
1.6.1: Quality studies using chemical method
− The first sample for chemical analysis of the Tuul river was collected in July, 1945 by
Russian research team [Tuvaanjav, 1983].
− Around 1950, a Russian researcher N.T.Kuznetsov was done a study for chemical component
of the Tuul River and concluded total mineralization not exceeds 200 mg l-1, hydro-carbonate,
sodium, potassium ions are predominant and chlorine, sulphate ions are minor [Tuvaanjav,
1972].
− Tuvaanjav, G in 1978, a researcher was done chemical analysis in the Tuul river and its
tributaries and found out the following conclusions; the Tuul river is belongs to hydro-
carbonate class and contains 78.9% calcium, 15.8% sodium, rest of percent is magnesium,
total mineralization is between 28.1-634.7 mg l-1. Upper reach of river and its tributaries
contain high amount of sulphate than chlorine, but contrary in lower reach. Because of, urban
area has negatively affects to river. Mineralization, pH and ammonia concentration increase
along the Selbe and the Uliastai Rivers [Tuvaanjav, 1978].
− In a research report written by Munguntsetseg, A and et all, they were estimated that the river
is totally self-purified along 170 km of downstream after wastewater from CWTP pours into
river [Munguntsetseg, et al., 1982].
− An article from Tuvaangav, G, the 40 million cubic meter water used in overall demand of
capital city and it strongly affects to natural source of water near by city [Tuvaanjav, 1983].
− A researcher Davaa, G in 1996 was done study in 12 points along the Tuul River and its 4
tributaries and noted that total mineralization increases along the river, because of, soil type
changes, precipitation decreases and evaporation increases [MNE, 1997a].
1.6.2: Quality studies using biological method
− By M. Kolikvitts and D.Marson in 1908-1909, they made the first attempt to assessing water
quality using characteristics of water insects that sensitive in living condition. After this
attempt, many scientists were trying to assess water quality using biological method and are
developing new methods in recent years [Batima, 1998].
− In 1930, the bacteriological study in river and spring water was done by A.Kharit, Russian
scientist. A scientist was trying to find relation between water bacteriology and stomach
illness [Adiyabadam, 1996].
− A research of aqua insects has done by I.M.Levanidova in 1947, a study about horsefly fauna
implemented by B.P.Belisheva, A.Dashdorj in 1958 and other aqua insects was studied by
Mongolian and Russian scientific team in 1972 and 1979, respectively. They estimated fauna
of aqua insects in surrounding area of the Ulaanbaatar city [Soninhishig, 1998].
− K.Shenon, E.A.Chebotarev and D.Lenat, biologists, were developed their methods for water
quality in 1948, 1986 and 1990, respectively [Zagas, 1998].
− A new method for water quality using water worm developed by Michigan and others in 1955
[Adiyabadam, 1996].
− In between 1990-1997, Institute of Biology and Institute of Geo-ecology of Mongolian
Academy of Sciences was cooperated with Institute of Water, MNE were done hydro-
biological research in the Tuul River and estimated 170 aqua insects. Moreover, team
assessed water quality using those fauna [Munguntsetseg, 1987].
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 9
− N.Soninhishig in 1998, she defended an MSc degree in Mongolia by thesis title “Algae in the
Tuul River”. She was exploring algae fauna in river and was assessed the following water
quality index by algae fauna [Soninhishig, 1998].
� Totally, 47 algae species found under Terelj Bridge of the Tuul River and 20 of them
live in natural pure water, 5 of them live in contaminated water; rest is not indicator
of water quality. She calculated quality index and river water belonged to first class.
Her conclusion was that river reach is pure.
� Entirely, 29 species found at confluence of the Khol and the Tuul rivers and nine of
them live in natural pure water, 5 of them live in contaminated water; rest is not
indicator of water quality. She calculated quality index, which belongs to the third
class and concluded that section of river is slightly polluted.
� Thirty sex species found under Bayanzurh Bridge, 13 were indicator of pure water, 7
were indicator of polluted water and concluded that; quality index is two, which
means, comparatively clean water.
� At the Songino (upper) study point was polluted because of quality index was four.
� At the Songino (down) stationary sampling point was heavily polluted, as a result of
quality index five.
� At the Chicken farm, river water assessed heavily polluted and quality index belonged
to fifth class.
� At the Khadanhyasaa and Altanbulag, stationary sampling points were dropped in
quality index four.
− An MSc thesis from B.Zagas and Ya.Oyunchuluun, both of them were used riverbed insects
for assessment of quality and concluded same results with previous study in 1998 and 1999,
respectively [Zagas, 1998], [Oyunchuluun, 1999].
− In 2000, Uranbileg, L set zones of river water quality in surrounding area of Ulaanbaatar
using biological method by Lenat, D, American scientist. She estimated totally four zones of
river water quality [Uranbileg, 2000].
Recent time, water quality is studying by several organization and researchers. Nowadays, there have
four institutes are concerning water issues, continuously, namely Central Laboratory of
Environmental Monitoring, Water sector of Institute of Geo-ecology, Water sector of National
Agency of Meteorology, Hydrology and Environmental Monitoring, Water Authority of Mongolia
and universities. In addition, number of projects are implementing in water issues and some of them
funded by Dutch government.
1.7: Research objectives
Research objective generated from practical problem in Mongolia.
− To assess river water quality using SWQ indices which based on nitrite NO2-, nitrate NO3
-,
ammonium NH4+, BOD5, COD-Mn and DO in case study of the Tuul River near by
Ulaanbaatar city
− To develop a prototype water quality model of the Tuul River using the Duflow Modelling
System
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 10
The following research tasks are developed from research objectives.
− Fieldwork, collecting samples, estimate location of point and non-point contamination sources
and collect existing hydraulic, hydro-chemical datasets
− Analyze water samples in laboratory
− Integrate the hydro-chemical data of selected rivers in last decade near to the Ulaanbaatar city
using availably data
− Assess river water quality using combination of nitrite, nitrate, ammonium, BOD5, COD-Mn,
DO and utilizing surface water quality indices
− Determine spatial and temporal changes of water quality
− Visualize thematic maps of river water contamination
− Develop a Tuul river flow model using DMS
− Build up a surface water quality model based on a flow model
1.8: Research questions
The following questions arose during the research and tried to find out answers:
− Which methods are suitable for our research? (standard methods or state-of-the-art)
− What kind of change was occurred in study period? (negative or positive)
− What type of model can we develop? (lumped, distributed and steady-state, continuous)
− Is DMS suitable for simulating flow and quality models of selected river? (yes or no)
− Is the output of research valuable and significant? (yes or no)
− If yes, that significance for whom? (local, regional and personal, groups)
1.9: Research phases
Entire period of this research divided in three main phases.
1) Pre-field work
− To review literatures
− To select methods and required data
− To prepare fieldwork
− To understand how Duflow modelling system works
2) During field work
− To collect samples and analyze in laboratory
− To gather availably datasets of last decade (1996 - 2006)
− To identify point and non-point sources of water contamination
− To collect a topographic map of study area
3) Post-field work
− To assess and classify river pollution using hydro-chemical datasets and SWQI
− To digitize topographical map of study area
− To estimate spatial and temporal changes of water quality
− To visualize thematic maps of river contamination
− To develop a river flow and a water quality model
− To set up a scenario for SWQ improvement in quality model
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 11
Table 7: Data, instrument requirement and sources
N Data Source
1 Literature ITC, internet, SCLM, Academic library of NUM and others
2 Sample from field
3 Dataset CLEM, NAMNEM and others
4 Map ALAGC
5 Documents (standard, book) MNE, NGIC project, MASM
6 Field equipment ITC
7 Computer ITC
8 Software ITC
9 Laboratory CLEM
1.10: Methodology
In this research, the following methodologies are selected, including:
− Literature review
This stage was encompassed the search and review of publications related with water quality
assessment, pollution mapping and water quality modelling issues in the study area and it is
important for estimate the research objectives.
− Field survey
This phase was involved measurement of chemical variables for water quality assessment of
the Tuul River and general survey of study area. The pH, EC, DO variables measured in the
field and rest of variables analyzed in the CLEM. Water samples are collected by standard
method MNS ISO 5667-6:2001 “Water quality. Sampling, 6th part. Guideline for sampling
from river and stream”. Therefore, contamination sources identified during field survey.
− Laboratory analysis
The chemical variables NO2-, NO3
-, NH4+ and COD-Mn were analyzed by spectrometer
method in the CLEM using HACH instrument. BOD5 was measured by dilution method in
same laboratory.
− Data collection
This step consisted of gathering the available data such as topographic map, hydro-chemical
datasets from previous years, hydraulic variable, physical parameter and climatic data from
different organization in Mongolia.
− Data analysis
In this stage, all data, which are collected and measured during fieldwork, assessed and
classified using SWQI
− Map visualization
Thematic maps for spatial and temporal changes of SWQ visualized using ILWIS software.
− Model development
A flow and a quality models were developed using Duflow Modelling System.
More than one method and technique were used to attain one task.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 12
Figure 4: Methodology flowchart
1.11: Research hypothesis
Obviously, the contamination of the Tuul River near by the Ulaanbaatar city has been continuously
increasing during last decade related to the increment of industrialization, agricultures and old,
insufficient sewage, treatment system.
It is quite possible to estimate temporal and spatial changes of water quality. Furthermore, according
to the User guide of Duflow software, it can develop a model of surface water quality using physical
parameter, hydrological variables, chemical variables and climatic data. Based upon that model, we
can set up a scenario and predict the effect of contamination source in water quality in the future.
1.12: Outline of the thesis
Background information of study field, the objectives of the research, recent problem statement,
previous studies in that field and methodology are included in the main subjects of chapter 1. Chapter
2 provides information about the selection of sampling site, data collection and field survey. The
following chapter 3, which is named water quality assessment, is mainly focused on spatial and
temporal assessment of surface water quality using geo-statistical tools and surface water quality
index. The thematic maps of surface water quality presented in chapter 3. Chapter 4 concentrated for
the Tuul River a flow model using Duflow modelling system. Process and results of a quality model
for DO in case of the Tuul River is mentioned in chapter 5. Chapter 6 provides the overall
conclusions, recommendations from this research and future studies.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 13
Chapter 2: Field survey and data collection
2.1: Field survey
The research field campaign was executed during 2 months, starting from 18 August until 24 October,
in the Ulaanbaatar city, Mongolia. Approximately, 800 km was travelled (including daily sample
transportation) during fieldwork and totally, 14 sample points were investigated for hydro-chemical
variables, 55 sites selected for measurement of general variables. HACH 40D field instruments and
ETREX hand-held GPS was used in the field campaign. Totally, 52 photos were taken in the field at
each site.
2.1.1: Selection of sampling site
According to classification of rivers based on discharge, drainage area and river width which
written in book by Chapman, D [WHO, 1996], the Tuul River is a small river. Because of, its
channel average width 35-75 meter during non-flooding time, the depth is 0.8-3.5 meter and the
velocity is 0.5-1.5 m s-1. The annual average discharge is approximately 26.6 m3 s-1. Drainage area
is in the region of 49,840 km2 and total length is 720 km, it flows generally in western direction.
Consistent with Mongolian river classification, developed by Davaa, G, which is based on long-
term annual mean flow, the Tuul River is a moderately big river.
CLEM chose the following 10 sampling points along the Tuul River and 4 sites in tributaries of the
Tuul River, one in Terelj, one in Uliastai, two in Selbe, in 1980th. Water quality and pollution in
surrounding area of the UB city has been monitored in those 14 points since 1980s.
During the field campaign, 14 samples collected from those sites for ammonium, nitrate, nitrite,
BOD5, COD-Mn examination and analyzed in the CLEM. In addition, some photos were captured
in the field. See some representative photos in figure 5 and Appendix 1.
Figure 5: Photo at Terelj-Terelj sampling point
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 14
The following a table and a map show the geographical locations of sampling points.
Table 8: Spatial and temporal information of water quality sampling points
N Name of sites Latitude Longitude Altitude
in m
Temporal
sampling Selection
1 Terelj - Terelj 47°59'30.67"N 107°27'35.55"E 1522 monthly Tributary of main river
2 Tuul - Uubulan 47°48'26.40"N 107°22'50.30"E 1383 monthly Base load
3 Tuul - Nalaih 47°49'14.00"N 107°15'56.40"E 1364 monthly Discharge from local WTS
4 Tuul - Bayanzurh 47°53'28.10"N 107°03'04.70"E 1309 monthly Inflow to city
5 Tuul - Zaisan 47°53'19.40"N 106°55'05.70"E 1293 monthly Centre of city
6 Tuul - Sonsgolon 47°52'28.70"N 106°46'50.10"E 1272 monthly Outflow from city
7 Tuul - Songino (upper) 47°51'17.80"N 106°41'23.20"E 1256 monthly Upper reach of CWTP
8 Tuul - Songino (down) 47°50'51.70"N 106°40'29.70"E 1254 monthly Lower reach of CWTP
9 Tuul – Chicken farm 47°46'21.00"N 106°35'59.20"E 1233 monthly Discharge from bio-industry
10 Tuul - Khadanhyasaa 47°45'08.90"N 106°30'02.60"E 1217 monthly Indicator of self-purification
and inflow to town
11 Tuul - Altanbulag 47°41'53.40"N 106°17'40.60"E 1182 monthly Indicator of self-purification
12 Uliastai - UB 47°54'07.80"N 107°01'51.77"E 1310 monthly Tributary of main river
13 Selbe - UB 47°54'30.77"N 106°55'55.77"E 1290 monthly Tributary of main river
14 Dund - UB 47°54'11.96"N 106°51'23.25"E 1276 monthly Tributary of main river
Hint: Geographical coordinate and altitude are measured by hand-held GPS, Etrex.
Figure 6: Locations of sampling points
There are five contamination point sources, which are represented by red triangles and 14 yellow
diamonds point out the sites where water samples were taken for above-mentioned five hydro-
chemical variables. In addition, pink dots are indicating the location of in-situ measurement places
where were measured general variables t0, pH, EC and DO using the HACH 40D field instrument for
water survey. Note: some of them might be coinciding.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 15
2.1.2: Sampling method
Surface water can be sampled using isokinetic and non-isokinetic samplers. Isokinetic samplers are
made such that stream water approaching and entering the sampler intake does not change in
velocity, and are consequently more expensive and harder to operate. In this research, we are used
different types of non-isokinetic samplers, namely open-mouth samplers. The hand-held plastic
bottle sampler is the simplest type of open-mouth sampler: a sample bottle is used as sampler. To
sample water from a river, hold the bottle mid of water depth, with the mouth facing slightly
downward. Turn the bottle upright to fill it and replace the bottle cap. [Dost, 2006].
2.1.3: Chemical analysis in field
A number of variables, pH, EC and DO must be measured in-situ very soon after the sample has
been collected because the value of these variables either change rapidly after sample collection
[Dost, 2006].
In the field, general variables t0, pH, EC and DO measured at least two times using HACH 40D a
field instrument. First, took water sample into a plastic container and measured by an instrument.
Secondly, probes immersed into flowing water and measured again. Difference of measurement
results was small and calculated average value.
Moreover, a field survey note prepared before the field campaign and totally, 55 sheets were filled
in-situ. Field survey sheet contains three clusters of information:
− Information of sampling location
− Sample information
− Measurement results
And, it includes the following information such as name of water body, basin, sub-basin,
geographical coordinate, elevation, land use around sampling point, possible contamination source,
name of sample collector, collected date, sample ID, sampling method, preservation method,
purpose of sampling, weather condition, any changes in weather condition, variable measured,
actual result, measured unit, analytical method and etc. For more information, see Appendix 2.
Beside of this, general variables were measured at 55 sites along the river. Aim of field
investigation was to estimate fluctuation of DO, pH, EC along river, identification of pollution
sources, and surveillance of entire study field. For the field measurements, see Appendix 3.
Brief steps of in-situ measurement:
− Prepare instrument
− Take a sample from river using a plastic container
− Measure variables
− Write and save results
− Immerse probes in river
− Measure variables, again
− Compare with previous result
− Calculate average number
− Fill field note
In situ measurement steps that are more detailed look at HACH 40D user manual or field guide for
water quality sampling and testing [Dost, 2006].
Chemical variables ammonium, nitrite, nitrate, COD-Mn and BOD5 in this study was measured in
the standard laboratory, CLEM in Ulaanbaatar, within a day after the sample had been collected
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 16
from above-mentioned sampling points. These hydro-chemical variables are unstable and
maximum permissible storage time is 48 hours.
In the field, 0.5-litre samples collected for laboratory analysis from river using plastic container,
were kept in a storage box and sent it to laboratory, analyzed by DR 2010 spectrophotometer
within a day. No preservative was added into samples. For more information about analysis
methods, see the procedure manual of the DR 2010 instrument.
2.2: Data collection
Beside of above-mentioned survey, some existing hydro-chemical, hydraulic and climatic datasets are
collected during field campaign. The following table has shown data collection from different
organization.
Table 9: Data collection and sources
N Data Implemented organization Used for
1 Hydro-chemical data CLEM Assessment, mapping, modelling
2 Hydraulic data WS, NAMHEM Modelling
3 Climatic data NAMHEM Modelling
4 Map and standards ALAGC and CSM Assessment, mapping, modelling
5 Pollution source CWTP and WSSA Assessment, modelling
2.2.1: Chemical data
The following chemical variables are included in hydro-chemical datasets:
− DO dissolved oxygen
− NH4+ ammonium
− NO2- nitrite
− NO3- nitrate
− COD-Mn chemical oxygen demand
− BOD5 biochemical oxygen demand
− T0water water temperature
Chemical dataset covers totally 11 years, 1996-2006. Water samples were taken from above-
mentioned 14 sampling points and examined in the CLEM. For an example of dataset, see
Appendix 4. All other hydro-chemical datasets prepared same as in Appendix 4.
Chemical dataset was used to assess quality of the Tuul River and its tributaries in surrounding
area of the UB city. Several quality distribution maps visualized by ILWIS based on assessment.
Moreover, DO, NH4+, BOD5 and water temperature data are input of a quality model.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 17
Figure 7: Monthly mean water temperature
0.00
2.00
4.00
6.00
8.0010.00
12.00
14.00
16.00
18.00
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05
Month
Wat
er te
mp
erat
ure
in C
2.2.2: Hydraulic data
Determining the hydrological regime of a water body is an important aspect of water quality
assessment. Discharge measurements, for example, are necessary for flow or mass balance
calculations and as inputs for water quality models [WHO, 1996].
Water Sector of NAMHEM has measured all of hydraulic variables such as velocity, discharge,
river width and depth, which were utilized this research. Hydraulic data, which was measured in
2005 and 2006, was used in river flow modelling. For hydraulic dataset, see Appendix 5.
Furthermore, daily discharge was gauged at Tuul-Ulaanbaatar and Tuul-Altanbulag stations,
respectively.
2.2.3: Climatic data
Climatic dataset includes hourly average solar radiation and monthly mean wind speed of the study
area. This dataset measured by NAMHEM at the Ulaanbaatar station, world weather station code
is 442920, and obtained from online open source IWEC. Downloaded wind speed and solar
radiation datasets are shown below in Figure 8 and Table 10;
Figure 8: Monthly average wind speed at Ulaanbaatar station
0
0.5
1
1.5
2
2.5
3
3.5
4
Month
Win
d sp
eed
in m
/s
Wind speed 1.4 1.8 3.2 2.9 3.7 3.6 3.2 3 2.8 3 2.6 1.6
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 18
Table 10: Hourly average solar radiation in W h/m2
Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0:01- 1:00 0 0 0 0 0 0 0 0 0 0 0 0
1:01- 2:00 0 0 0 0 0 0 0 0 0 0 0 0
2:01- 3:00 0 0 0 0 0 0 0 0 0 0 0 0
3:01- 4:00 0 0 0 0 0 0 0 0 0 0 0 0
4:01- 5:00 0 0 0 0 0 0 0 0 0 0 0 0
5:01- 6:00 0 0 0 0 0 11 0 0 0 0 0 0
6:01- 7:00 0 0 0 69 135 155 103 33 0 0 0 0
7:01- 8:00 0 0 63 282 353 329 210 177 127 6 0 0
8:01- 9:00 0 84 287 542 472 422 318 312 304 172 61 0
9:01-10:00 192 325 475 704 479 461 385 382 420 384 243 160
10:01-11:00 367 543 523 627 468 460 411 398 456 502 399 381
11:01-12:00 517 594 585 600 477 466 389 382 468 555 488 488
12:01-13:00 541 564 596 552 461 464 344 346 441 558 520 492
13:01-14:00 534 513 567 441 432 452 290 308 408 522 513 449
14:01-15:00 509 514 560 502 405 439 290 348 430 488 465 434
15:01-16:00 423 481 531 576 366 412 283 379 415 395 334 316
16:01-17:00 203 362 458 442 330 375 270 379 354 225 95 38
17:01-18:00 0 140 354 463 307 339 276 322 258 39 0 0
18:01-19:00 0 0 55 310 210 264 219 203 54 0 0 0
19:01-20:00 0 0 0 6 48 128 86 19 0 0 0 0
20:01-21:00 0 0 0 0 0 0 0 0 0 0 0 0
21:01-22:00 0 0 0 0 0 0 0 0 0 0 0 0
22:01-23:00 0 0 0 0 0 0 0 0 0 0 0 0
23:01-24:00 0 0 0 0 0 0 0 0 0 0 0 0
Source: [ASHRAE, 2001]
2.2.4: Map and standards
Topographic map with scale 1:200000 were digitized and used for the spatial and temporal SWQ
distribution maps. Printed-paper maps were obtained from ALAGC.
The following two Mongolian National standards were used for the water sample collection and
SWQ assessment;
− MNS ISO 5667-6:2001 “Water quality. Sampling, 6th part. Guideline for sampling from
river and stream”
− MNS 4586:98 “Surface water quality. Permissible level of surface water variables”
Those standards were collected from CSM.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 19
Table 11: Permissible level of surface water variables
Variable name Chemical formula Unit Permissible level
Hydrogen ion activity pH 6.5 – 8.5
Dissolved oxygen O2 mg l-1 6 and 4
Biochemical oxygen demand - 5 day BOD5 mg l-1 3
Chemical oxygen demand - Manganese COD-Mn mg l-1 10
Ammonium NH4-N mg l-1 0.5
Nitrite NO2-N mg l-1 0.02
Nitrate NO3-N mg l-1 9.0
Phosphorus PO4-P mg l-1 0.1
Chloride CI- mg l-1 300
Fluoride F mg l-1 1.5
Sulphate SO4-2 mg l-1 100
Manganese Mn mg l-1 0.1
Nickel Ni mg l-1 0.01
Copper Cu mg l-1 0.01
molybdenum Mo mg l-1 0.25
Cadmium Cd mg l-1 0.005
Cobalt Co mg l-1 0.01
Lead Pb mg l-1 0.01
Arsenic As mg l-1 0.01
Total Chromium Cr mg l-1 0.05
Chromium with hexavalence Cr+6 mg l-1 0.01
Zinc Zn mg l-1 0.01
Mercury Hg mg l-1 0.1
Hint: DO is 6 in warm session and 4 in cold session (water covered by ice)
Source: [CSM, 1998]
2.2.5: Pollution source
Chemical monthly data of discharge from the CWTP, the largest pollution point source, were
provided and supported by the NGIC project. The CWTP is responsible for the collection and
treatment of industrial and domestic sewage water in the Ulaanbaatar city. First mechanical
treatment filter installed in 1969. Overall treatment capacity is 230,000 m3 day-1. Recently,
150,000-160,000 m3 day-1 domestic and industrial sewage water is passing through a treatment
plant. In CWTP laboratory, 45 chemical and biological analyses can do. From those, 15 analysis
methods meet ISO standards.
Most recent problem is the electricity supply. If electrical power cut off during an hour, 4,500 m3
wastewater flows into the Tuul River without any treatment [CWTP, 2006]. For the dataset see
Appendix 6.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 21
Chapter 3: Water quality assessment
3.1: Method for surface water quality assessment
In Mongolia, surface water quality is estimated by two different methods;
1. Water quality grade developed by Water sector of MNE (former name)
2. Water quality index developed by Erdenebayar,Ya., and Bulgan,T
In this research, second method was used to evaluate the surface water quality.
The water quality index is estimated using the following formulae:
Equation 1: Surface water quality index
n
Pl
C
W i
i
i
qi
Σ
=
Where:
Wqi water quality index
Pli permissible level of i-th variable
n number of variables
Ci concentration of i-th variable
The Mongolian National Standard MNS 4586-98, which was developed by CSM in 1998, determine
the maximum permissible levels of chemical variables in the surface water. Permissible levels (see
table 11 in chapter 2), DO, BOD5, COD-Mn, ammonium, nitrate, nitrite and other variables should be
included in the calculation of quality index [Davaa, et al., 2006].
After calculation of quality index, the water sample is classified using this system (Table 12):
Table 12: Assessment of surface water quality
Water quality Quality index
degree class
< 0.30 1 Very clean
0.31 – 0.89 2 Clean
0.90 – 2.49 3 Slightly polluted
2.50 – 3.99 4 Moderately polluted
4.00 – 5.99 5 Heavily polluted
6.00 < 6 Dirty
Source: [MNE, 2006]
Surface water usage depends on quality of the water. The MNE of Mongolia determines the definition
of the WQ classes and water usage of specific waters.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 22
Table 13: Definition for classification of surface water quality in Mongolia
Classes Definition
1
Very clean
Sections of water bodies with pure, usually oxygen saturated, nutrient-poor water; low
bacteria content and directly use for drinking purpose, suitable for all kinds of water
usage.
2
Clean
Water bodies with small amount of inputs of organic or inorganic nutrients but without
or slightly oxygen depletion, low bacteria content and use for drinking and food
production purposes after disinfection and filtration, directly utilize for fishing factory.
3
Slightly
polluted
Section of water bodies with slight pollution, not a good oxygen supply, inputs of
organic or inorganic nutrients, some bacteria content and not suitable for drinking and
food production purposes, if no choice use it above mentioned purposes after treatment,
disinfection, filtration and can use directly for livestock, recreation, sport purposes.
4
Moderately
polluted
Water bodies with inputs of organic, oxygen consuming substances capable of producing
critical oxygen depletion; fish kills possible during short periods of oxygen deficiency;
declining numbers of macro-organisms; certain species tend to produce massive
populations and use for irrigation, industrial process after filtration.
5
Heavily
polluted
Sections of water bodies with heavy organic, usually low oxygen content; localised
deposits of anoxic sediment; filamentous sewage bacteria, occasional mass development
of a few micro-organisms, which are not sensitive to oxygen deficiency, periodic fish
kills occur, after filtration, use for some industrial process, which is not take part in
human.
6
Dirty
If value of water quality index exceeds 5th degree, it is belongs to this class. Sections of
water bodies with excessive pollution by organic, oxygen depleting sewage; processes of
putrefaction predominate; prolonged periods of very low oxygen concentrations or total
deoxygenating; mainly colonised by bacteria, no fish stocks; loss of biological life in the
presence of severe toxic inputs and can not use any purpose.
Source: [MNE, 1997b]
3.2: Basic geo-statistical analysis
The basic geo-statistical analysis was done by specific software “R” version 2.5.1, which developed in
2001 and recently updated in 27.06.2007. Chemical dataset of the Tuul River consists of totally 11
years data that were measured by CLEM between 1996-2006. Chemical variables, DO, BOD5, COD-
Mn, NH4+, NO2
-, NO3-, used to calculate water quality index.
3.2.1: Descriptive statistics
The Tuul River data frame contains two coordinates (here named E (UTM) and N (UTM)), two
categorical variable (name of sampling point) and ID. Four continuous variables, representing
water quality of the Tuul River expressed by index, time steps starting from 1996 until 2006 and
the distance measured from the upper reach by meter, water quality classified in 6 different
degrees. Totally, 1192 observations were done at 14 sampling points along the Tuul River and its
tributaries between 1996 and 2006.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 23
Figure 9: Data frame of the Tuul River dataset
Table 14: Statistical summary of water quality index
Minimum 1st quartile Median Mean 3rd quartile Maximum
0.15 0.33 0.58 1.68 1.45 34.2
The minimum and maximum values of WQI are 0.15 and 34.2, respectively. The difference
between minimum and maximum value is 34.05.
Figure 10: Highest value of WQI
The highest value of WQI is 34.2 in row 852. This attribute value measured in December 2004, a
sample took at Tuul-Songino (down), which is located in 625304 East, 5300732 North and
distance from upper reach is 61017 m. In addition, almost all high values were measured at
sampling points in downstream, after the discharge from the CWTP enters the river.
According to the classification of surface water pollution in Mongolia, the first critical value in
river water is 0.3 and between 0.31 and 0.89 surface water concerns second level of water quality,
namely clean. Equal to or greater than 0.9 until 2.49, it belongs to third level and from 2.5 to 3.99
is fourth level. Fifth level has value between 4.0 and 5.99. Greater than or equal to 6.0, it concerns
sixth level and that water is strictly forbidden to be used for any purpose.
Number of observations with threshold values and its percentage in the entire dataset can be
estimate using the logical expressions in the R software.
Table 15: The quantity of samples with critical values
Threshold values Number of observations Percentage in total observations
<= 0.30 244 20.5
>= 0.31 948 79.5
>= 0.90 453 38.0
>= 2.50 167 14.0
>= 4.00 97 8.1
>= 6.00 74 6.2
Total numbers of observations, less than and equal to 0.3, are 244 and represent 20.5 percent of the
total of 1192 observations. The numbers of observations, which greater than and equal to 6.0, are
74 and get 6.2 percent in total observations. Cumulated observations 948 catch 79.5% that belongs
to observation with greater than and equal to 0.31.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 24
3.2.2: Exploratory graphics
The R geo-statistical software package provides a rich environment for statistical visualization.
There are two graphics systems: the base system (in the graphics package, loaded by default when
R starts) and the Trellis system (implemented in R by the lattice package) [Rossiter, 2007]. The
subsequent figures visualized by the base system.
Figure 11: Histogram and box plot of water quality index
Histograms have one primary deficiency – their visual impression depends on the number of
categories selected for the plot. Comparisons of shape and similarity of histograms of the same
data depend on the choice of bar widths and centre [Helsel and Hirsch, 1992].
This histogram visualizes the frequency distribution of water quality index. The distribution is
strongly right-skewed and high values are very rare. The modal value is in the 0 to 2 range.
Another useful univariate plot is the box plot. Box plot which gives a rough idea of the shape of a
unimodal distribution [Rossiter, 2007]. Box plot in Figure 11 visualizes the quartiles distribution
of the water quality index. Most of attribute values are between 0 and 2.
Figure 12: Box plot for spatial variability of WQI
The second box plot shows WQ variability at
sampling points along the Tuul River. Most
dynamic one is a sampling point 11, namely
Tuul-Songino (down). At this point, WQ is
heavily influenced by treated water discharges
from the CWTP. That means, the variability of
WQI is directly related to human activity. Then,
it is naturally purified along the river in
downstream direction. First 10 sampling points
have less variability of WQI due to lesser
human impact (except some tourist camps and
towns) to river water quality.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 25
3.3: Spatial water quality assessment
At a given river station water quality depends on many factors, including: (i) the proportion of surface
run-off and groundwater, (ii) reactions within the river system governed by internal processes, (iii) the
mixing of water from tributaries of different quality (in the case of heterogeneous river basins), and
(iv) inputs of pollutants [WHO, 1996].
Spatially, river water quality decreases along the river in downstream direction. Several point and
non-point contamination sources exist in study area. The point sources of pollution in the Tuul River
are improper treated wastewater from (i) Nalaih (1,400 m3 day-1), (ii) Niseh (400 m3 day-1), (iii)
CWTP (190,000 m3 day-1), (iv) Bio-industry (490 m3 day-1) and (v) Bio-Songino wastewater treatment
station (600 m3 day-1). The largest point source is CWTP, which is situated in western side of UB.
Daily mean discharge from that point is approximately 190,000 m3 day-1. The CWTP is responsible
for the collection and treatment of industrial and domestic sewage in the Ulaanbaatar city. Main non-
point source is the UB city that produces 800-1,000 tons of dry wastes per day [Orchlon, 1995].
Water pollution of the Tuul River basin is continuously increasing from upper to its lower reach.
Naturally, upstream of the Tuul River is running through mountainously area and there has high
velocity and turbulent. Hence, upper part of river has more oxidization potential, re-aeration and self-
purification. When it comes to the Ulaanbaatar city, natural conditions change from mountainous
region to valley. In the valley, velocity and turbulence of river decreases, then capability of
oxidization, re-aeration and self-purification also reduces as well. This is the natural factor of
possibility to store contaminant elements in the river water a longer time and distance [Altansukh,
2000].
Scatter plots were produced to illustrate the spatial distributions of the WQI.
Figure 13: Correlation between distance & WQI
Until around 60 km from upper reach, the water
quality is almost stable (comparing with high peak
value) which means did not reach to highest critical
value. Nevertheless, high peaks are starting from 11th
sampling point. Because of, that reach is joint part of
the Tuul River and discharge from the CWTP.
Based on above analysis, entire Tuul River dataset
separated into two sub-datasets, which are upstream
and downstream. Reason of this, the Tuul River
dataset consists of two different populations,
statistically. The upstream dataset contains chemical
analysis data of the Tuul River water quality from
upstream (sampling point number 1) until sampling
point number 10 which is located just upper reach of effluent from the CWTP. And, the downstream
dataset contains data from sampling point number 11 that is located lower reach of runoff from the
CWTP till last sampling point number 14 of this study.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 26
Figure 14: Correlation between distance and WQI in the upstream section
0123456789
0 10000 20000 30000 40000 50000 60000
Distance, m
Ind
ex
In the upstream reach, there is little or no correlation between distance and water quality (nearly
stable). Fluctuation of water quality slightly changes along the river. Moreover, the quality index does
not reach the maximum critical value of 6 (ignoring some outliers). Two point pollution sources out of
five operate in upstream study area, namely Nalaih and Niseh WTSs. Total amount of effluents,
which are released from those two points, are approximately 1,800 m3 day-1. This amount of discharge
does not have a strong affect on the river (annual average flow 20 m3 s-1 > 0.021 m3 s-1). In addition,
distance between two point sources is around 54 km along the river. This is an adequate amount of
distance for river self-purification after the first effluents from Naliah WTS pour into the river water.
Figure 15: Water quality fluctuation along the upstream in 1996-2006
0
1
2
3
4
5
6
Ter
elj -
Ter
elj
Tuu
l -U
ubul
an
Tuu
l - N
alai
h
Tuu
l -B
ayan
zurh
Ulia
stai
- U
B
Tuu
l - Z
aisa
n
Tuu
l -S
onsg
olon
Tuu
l -S
ongi
no(u
pper
)
Name of sampling point
Inde
x
In the downstream reach, from a main pollutant source onwards, the quality index inversely correlates
to distance. The values in Figure 16 show that the index already exceeds the maximum critical value.
Because of, that is largest point source of contamination. Three point sources are situated in
downstream. Total volume of discharges from the CWTP, Bio-industry and Bio-Songino WTSs are
approximately 191,090 m3 day-1 and distance between point sources is around 2.5 km. This distance is
not sufficient for the self-purification process to take fully place, especially after large volumes of
effluent from CWTP pour into the river (annual average flow 20 m3 s-1 > 2.2 m3 s-1).
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 27
Figure 16: Correlation between distance and WQI in the downstream section
0
10
20
30
40
60000 65000 70000 75000 80000 85000 90000 95000
Distance, m
Ind
ex
When effluent from any source joins to the river, physical, biological and chemical processes such as
dilution, dispersion, advection and re-aeration occur in the river water, simultaneously. Those
processes are very beneficial in terms of water quality. Because river water has a self-purification
capacity, as a result of those processes. Self-purification is spatial and temporal process and depends
on many factors such as water flow velocity, river meandering, volume of fluid, chemical contents
and concentration, dilution, soil characteristics, diffusion, water turbulence, advection, temperature,
re-aeration and etc. Pollution of the Tuul River reduces along the downstream direction, but it is not
yet completely purified after a downstream distance of 35 km (after the main outfall).
3.4: Temporal water quality assessment
Water quality and pollution in surrounding area of the UB city was monitored in 14 stationary points
since 1980s. Ten sampling points along the Tuul River and 4 points in tributaries of the Tuul River, 1
in Terelj, 1 in Uliastai, 2 in Selbe, were chosen by CLEM in 1980th. In the last decade, fast
urbanization in combination with increasing number of tourist camps, agricultural and mining
activities have had a significant negative impact on the Tuul River’s quality and its associated
ecosystems. Consequently, the water became seriously polluted, lost its clarity and transparency and
its self-purification distance and time increased in year by year [Baasandorj and Davaa, 2006].
Figure 17: Correlation between time & WQI (1996-2006)
0
5
10
15
20
25
30
35
40
Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06
Time steps
Inde
x
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 28
Therefore, the general trend of water quality index and variability is increasing throughout the study
period. In the year 1999, the water quality was most stable and a year with high variability is 2005.
Reason of that is a new filtering system installed in the CWTP in 1999 by JICA support. Moreover,
this system is not renewed until now. In addition, amount of wastewater that flows through the
CWTP, is continuously increasing due to population growth of the capital city Ulaanbaatar and
industrialization.
Figure 18: Water quality fluctuation at Tuul-Uubulan SP during study period
0.34 0.32 0.29 0.32 0.30 0.330.28
0.350.32
0.43 0.41
0.00
0.10
0.20
0.30
0.40
0.50
1996 1998 2000 2002 2004 2006
Time step
Inde
x
Generally, the water quality did not snow large changes at selected sampling points in the upstream
reach during the study period. No large contamination sources of influence exist yet in the upper
reaches.
Figure 19: Water quality fluctuation at Tuul-Songino (upper) SP during study period
0.46
0.34
0.49
0.35 0.38 0.35
0.47
0.56 0.530.46
0.39
0.00
0.10
0.20
0.30
0.40
0.50
0.60
1996 1998 2000 2002 2004 2006
Time step
Inde
x
However, the water quality shows a slightly increment of general trend in upstream. In the
downstream section, water quality decreased during the study time and the general trend was an
increase in the WQI.
Figure 20: Water quality fluctuation at Tuul-Songino (down) SP during study time
3.014.33 3.93
1.44
2.904.12
7.77
5.926.74
7.458.60
0.00
2.00
4.00
6.00
8.00
10.00
1996 1998 2000 2002 2004 2006
Time step
Inde
x
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 29
The following time series of thematic WQ maps show the temporal changes of water quality along the
Tuul River in 3 selected years.
Figure 21: Water quality along the Tuul River in 1996
In the 1996, the Tuul River was not seriously polluted, yet. Classes of heavily polluted and dirty water
are not visible in the map.
Figure 22: Water quality along the Tuul River in 2002
In the 2002, the Tuul River starts getting seriously polluted. Classes of heavily polluted and dirty
water can be visualized in the map. Because of, rapid urbanization and increase of industrial activities,
including frequent spills from industries and insufficient operation of the CWTP.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 30
Figure 23: Water quality along the Tuul River in 2006
In the 2006, the Tuul River starts
getting strongly polluted. Because
of, the efficiency of the CWTP
operation began to fail due to poor
maintenance, lack of spare parts,
outdated equipment, frequent
power shortages and plus above-
mentioned reasons.
3.4.1: Seasonal water quality assessment
In fact, Mongolia has four seasons. In this study, all months of the year were divided into two
periods, warm and cold. Cold period continues from November until end of March and average air
temperature is below zero and deposits snow. This natural phenomena negatively affects the
internal processes in the river and the interaction between the river and other natural components.
That means, self-purification process cannot takes place in this period. Because of, river discharge
reaches between 0 - 2 m3 s-1 and discharge from the CWTP is 2.2 m3 s-1, normally. Warm period
continues from April until end of October and mean air temperature is above zero, with rainfall.
Figure 24: Seasonal water quality in upstream reach
0.00
0.20
0.40
0.60
0.80
1.00
Tuul-Uubulan Tuul-Nalaih Tuul-Bayanzurh
Tuul-Zaisan Tuul-Sonsgolon
Tuul-Songino(upper)
Sampling point
Index
Overall mean in warm period Overall mean in cold period
In the upstream section, there is no big pollutant source (ignoring effluent from Nalaih WTS of
approximately 0.02 m3 s-1). That is why, the water quality index calculated with small values and
difference between indices in warm and cold period is little.
Figure 25: Seasonal water quality in downstream reach
0.00
2.00
4.00
6.00
8.00
10.00
12.00
Tuul-Songino (down) Tuul-Chicken farm Tuul-Khadanhyasaa Tuul-Altanbulag
Sampling point
Index
Overall mean in warm period Overall mean in cold period
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 31
In downstream section, several pollutant point sources exist and total discharge is 2.2 m3 s-1,
approximately. Water quality index calculated with high values in cold period and difference
between warm and cold period’s indices is high. Because of, self-purification process intensively
takes place in warm period and this process is almost interrupted during the cold period. That is
why, overall mean index of water quality calculated maximum value with 2.94 in warm period and
11.05 in cold period. See Table 16.
Table 16: Annual mean water quality index
Date Tuul-
Uubulan Tuul-Nalaih
Tuul-Bayanzurh
Tuul-Zaisan
Tuul-Sonsgolon
Tuul-Songino (upper)
Tuul-Songino (down)
Tuul-Chicken
farm
Tuul-Khadanhyasaa
Tuul-Altanbulag
Warm period Average-96 0.40 0.42 0.46 0.39 0.56 0.56 1.13 1.84 1.75 1.42 Average-97 0.35 0.32 0.41 0.35 0.40 0.37 1.56 1.87 1.65 1.44 Average-98 0.36 0.36 0.48 0.35 0.38 0.48 3.26 2.73 2.56 2.01 Average-99 0.36 0.40 0.40 0.36 0.59 0.36 0.91 1.36 1.50 1.36 Average-00 0.36 0.36 0.34 0.37 0.34 0.38 2.27 2.48 3.11 1.55 Average-01 0.34 0.38 0.41 0.38 0.37 0.36 2.36 2.11 1.98 1.37 Average-02 0.32 0.38 0.33 0.63 0.55 0.54 4.26 3.31 2.65 2.18 Average-03 0.41 0.47 0.45 0.44 0.46 0.60 4.55 6.52 1.33 2.69 Average-04 0.39 0.51 0.48 0.54 0.48 0.58 1.27 1.68 1.96 1.58 Average-05 0.37 0.50 0.39 0.39 0.34 0.39 3.69 3.62 2.43 1.70 Average-06 0.36 0.47 0.38 0.34 0.41 0.41 6.87 4.85 2.54 2.00
Overall mean 0.37 0.42 0.41 0.41 0.44 0.46 2.92 2.94 2.13 1.75 Cold period
Average-96 0.40 1.06 0.40 0.30 0.35 7.47 7.59 2.00 1.26 Average-97 0.27 0.25 0.24 0.51 2.53 0.30 11.19 8.69 3.40 1.58 Average-98 0.24 0.29 0.27 0.30 0.30 0.26 3.65 6.09 1.82 0.77 Average-99 0.27 0.32 0.28 0.26 0.28 3.16 3.65 3.02 1.89 Average-00 0.28 3.44 0.52 7.08 3.53 1.51 0.68 Average-01 0.37 0.50 0.32 0.43 0.34 0.26 8.56 4.19 1.92 1.42 Average-02 0.28 0.50 0.35 0.29 15.20 10.52 4.11 3.17 Average-03 3.42 0.85 0.32 0.39 12.10 5.52 2.57 Average-04 0.55 0.39 0.25 0.17 0.17 17.78 12.62 5.21 3.25 Average-05 2.05 0.73 0.48 1.05 1.15 0.87 17.76 13.40 9.36 3.01 Average-06 0.65 0.62 0.35 0.23 0.21 0.23 17.57 13.07 5.59 20.80
Overall mean 0.80 0.81 0.33 0.42 0.65 0.36 11.05 8.08 3.68 3.78
3.5: Chapter conclusions
The following conclusion can be drawn from this chapter:
− The highest value of the WQI observed was 34.2. This attribute value was measured in
December 2004, in a water sample from Tuul-Songino (down).
− Statistically, the entire dataset consisted by two different populations, natural water and
natural water mixed with discharge from the CWTP.
− Until the main pollutant source outfall enters the Tuul River, there is no significant
correlation between distance and water quality. However, from the main pollutant source in
downstream, the water quality inversely correlates to distance.
− The general trend of water quality index and variability increases throughout the time steps.
In the year 1999, the water quality was most stable and a year with high variability is 2005.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 32
− Pollution of the Tuul River reduces as a function of distance in the downstream section, but
completely self-purification was not observed even at the last sampling point of this study (35
km downstream major outfall).
− The natural phenomenon of cold period (t < 00C), negatively affects the self-purification
process of the Tuul river. Reason is also that the discharge from pollutant sources is in this
period greater than the natural river discharge.
− The CWTP remains the biggest pollution source on this section of the Tuul River.
3.6: Limitation in WQA
In this section, the correlation between water quality and quantity could not be studied in detail. One
reason is that the Tuul River hydraulic parameters measurement and quality analysis are implemented
by two different organizations. Moreover, measurement dates and locations of both organizations are
not coinciding. For example, daily discharge measurement are done in three different stations, which
belong to WS, NAMHEM, situated in study area. Water quality analysis, implemented by CLEM
once a month, is done in different locations from the hydraulic measurements. Furthermore, quality
analysis date is not written in the chemical dataset that provided by CLEM.
However, monthly mean discharge value and monthly WQI at Tuul-Zaisan sampling point were
plotted in the following graph.
Figure 26: Correlation between river monthly mean discharge and monthly WQI
Correlation between river discharge and WQ(Tuul-Zaisan)
0.1
0.3
0.5
0.7
0.9
1.1
Ap
r-9
6
Ap
r-9
7
Ap
r-9
8
Ap
r-9
9
Ap
r-0
0
Ap
r-0
1
Ap
r-0
2
Ap
r-0
3
Date
WQ
I
0.00
20.00
40.00
60.00
80.00
100.00
Q, m
3 s
-1
WQI Q
Obviously, water quality can be related to quantity. In general, when river discharge increases then
WQI decreases (only case of effluent from pollution source is constant). However, we can observe
hysteresis effects and large variations and deviations can be observed.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 33
Chapter 4: River flow model
4.1: General introduction of flow model
The Duflow is a linked flow and water quality modelling package, which can be used to simulate one
dimensional flow and water quality in a network of open water courses [DMS, 2004b].
The Duflow is a numerical model. To model the water system network, the real water system has to
be sub-divided into sections and nodal points. Sections reflect river reaches of uniform characteristics
and nodal points are used to connect the sections. For each section characteristics, dimensions, like
length and cross sectional profile, have to be defined. Manning or Chezy coefficients can be used to
account for section roughness [Makkinga, et al., 1998]. Within the network, several types of
structures can be defined (culverts, overflows, underflows, siphons and pumps). The flow model also
includes a simple rainfall-runoff module. At each nodal point, a catchment area can be defined and
using a runoff coefficient, rain losses can be taken into account. Furthermore, at each nodal point
additional flows can be defined to account for discharges and tributaries not included in the network.
At the system boundaries additional flows, discharge curve or water levels can be used as boundary
conditions. All boundary conditions and discharges can be entered as constants or as time functions
[Makkinga, et al., 1998].
4.2: Background theory of flow simulation
The following equations mainly used to simulation and evaluation of a flow model.
4.2.1: Saint-Venant equation
DUFLOW simulation is based on the one-dimensional partial differential equation (Saint-Venant)
that describes non-stationary flow in open channels (Abbott, 1979; Dronkers, 1964).
Several assumptions consider for St. Venant equations:
− Flow is one-dimensional
− Hydrostatic pressure prevails and vertical accelerations are negligible
− Streamline curvature is small
− Bottom slope of the channel is small
− Manning equation is used to describe resistance effects
− The fluid is incompressible [Merwade, 2005]
These equations, which are the mathematical translation of the laws of conservation of mass and of
momentum:
Equation 2: Mass equation
0x
Q
t
B =∂∂+
∂∂
Equation 3: Momentum equation
( ) ( )φ−Φγ=+∂α∂+
∂∂+
∂∂
coswaARC
QQg
x
Qv
x
HgA
t
Q 22
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 34
Equation 4: Discharge equation A*vQ =
Where:
t time [s]
x distance as measured along the channel axis [m]
H(x, t) water level with respect to reference level [m]
v(x, t) mean velocity (averaged over the cross-sectional area) [m s-1]
Q(x, t) discharge at location x and at time t [m3 s-1]
R(x, H) hydraulic radius of cross-section [m]
a(x, H) cross-sectional flow width [m]
A(x, H) cross-sectional flow area [m2]
b(x, H) cross-sectional storage width [m]
B(x, H) cross-sectional storage area [m2]
g acceleration due to gravity [m s-2]
C(x, H) coefficient of De Chézy [m1/2 s-1]
w(t) wind velocity [m s-1]
Φ(t) wind direction [degrees]
φ(x) direction of channel axis, measured clockwise from the north [degrees]
γ(x) wind conversion coefficient [-]
α correction factor for non-uniformity of the velocity distribution in the advection
term, defined as:
Equation 5: Correction factor for non-uniformity of velocity distribution
( )∫=α dydzz,yvQ
A 2
2
A cross-section [m2]
The mass equation states, if the water level changes at some location this will be the net result of
local inflow minus outflow. The momentum equation expresses that the net change of momentum
is the result of interior and exterior forces like friction, wind and gravity. For the derivation of
these equations, it has been assumed that the fluid is well mixed and hence the density may be
considered constant [DMS, 2004b].
4.2.2: Manning equation
The Manning equation is the most commonly used equation to analyze open channel flows. It is a
semi-empirical equation for simulating water flows in channels and culverts where the water is
open to the atmosphere, i.e. not flowing under pressure, and was first presented in 1889 by Robert
Manning. The channel can be any shape - circular, rectangular, triangular, etc. The units in the
Manning equation appear to be inconsistent; however, the value k has hidden units in it to make
the equation consistent [Edwards, 1998].
Equation 6: Hydraulic radius
P
AR =
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 35
Equation 7: Slope of riverbed
L
hS f∆
=
Equation 8: River velocity
2
1
3
2
SRn
kV =
Where:
V velocity, m s-1
A cross-sectional area, m2
n Manning coefficient
P wetted perimeter, m
R hydraulic radius, m
S slope
hf difference between bottom of channels, m
L length of channel, m
k 1.0 for unit conversion
Slope of selected section of the Tuul River is approximately 0.00198 or 0.2%. Reason of calculate
the slope, if slope exceeds 2% then DMS cannot handle it.
4.2.3: Froude number
The Froude number, Fr, is a dimensionless value that describes different flow regimes of open
channel flow. The Froude number is a ratio of inertial and gravitational forces.
− Gravity (numerator) - moves water downhill
− Inertia (denominator) - reflects its willingness to do so [White, 1998].
Equation 9: Froude number
gD
VFr =
Where:
D hydraulic depth, m
g gravity, m s-1
A cross-sectional area, m2
b flow width, m
Equation 10: Hydraulic depth
b
AD =
When:
Fr = 1, critical flow,
Fr > 1, supercritical flow (fast rapid flow),
Fr < 1, subcritical flow (slow / tranquil flow)
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 36
The Froude number is a measurement of bulk flow characteristics such as waves, bed forms, and
flow, depth interactions at a cross section or between boulders. The denominator represents the
speed of a small wave on the water surface relative to the speed of the water, called wave celerity.
At critical flow celerity equals flow velocity. Any disturbance to the surface will remain stationary.
In subcritical flow the flow is controlled from a downstream point and information is transmitted
upstream. This condition leads to backwater effects. Supercritical flow is controlled upstream and
disturbances are transmitted downstream [Furniss, et al., 2006].
At Ulaanbaatar station, Fr is equal to 0.74, which means subcritical flow.
74.0m63.0*s/m81.9
s/m85.1
gD
VFr
2===
At Altanbulag station, Fr is equal to 0.45, which means subcritical flow.
45.0m67.0*s/m81.9
s/m16.1
gD
VFr
2===
Maximum velocity and associated width, hydraulic area in study period are used to calculate
Froude number at specific location. Reason of Fr calculation, if flow is supercritical then DMS
cannot simulate the river flow.
4.2.4: Nash-Sutcliffe coefficient
The Nash-Sutcliffe model efficiency coefficient is used to assess the predictive power of
hydrological models. It is defined as:
Equation 11: Nash-Sutcliffe coefficient
( )( )∑
∑
=
=
−
−−=
T
1t
2
oto
T
1t
2tm
to
QQ1E
Where:
Qo observed discharge
Qm modelled discharge
Qt discharge at time t
Nash-Sutcliffe efficiencies can range from -∞ to 1. An efficiency of 1 (E = 1) corresponds to a
perfect match of modelled discharge to the observed data. An efficiency of 0 (E = 0) indicates that
the model predictions are as accurate as the mean of the observed data, whereas an efficiency less
than zero (-∞<E<0) occurs when the observed mean is a better predictor than the model.
Essentially, the closer the model efficiency is to 1, the more accurate the model is [Nash and
Sutcliffe, 1970].
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 37
4.2.5: Statistical methods ME, MAE, RMSE
Mean error measures the average between observed and calculated variables. Both negative and
positive differences are incorporated in the calculation; they may cancel out the error. As such
small error may not indicate a good calibration [Rientjes, 2006].
Equation 12: Mean error
( )∑ −−=
n
1ico
n
1ME
Where:
n number of values
o observed values
c calculated values
The mean absolute error measures the average magnitude of the errors in a set of forecasts, without
considering their direction. It measures accuracy for continuous variables. The MAE is the average
over the verification sample of the absolute values of the differences between forecast and the
corresponding observation. The MAE is a linear score, which means that all the individual
differences are weighted equally in the average.
Equation 13: Mean absolute error
∑ −−=
n
1ico
n
1MAE
The root mean squared error measures the difference between forecast and corresponding observed
values are each squared and then averaged over the sample. Finally, the square root of the average
is taken. Since the errors are squared before they are averaged, the RMSE gives a relatively high
weight to large errors. This means the RMSE is most useful when large errors are particularly
undesirable.
Equation 14: Root means squared error
( )5.0
n
1i
2ico
n
1RMSE
−= ∑ −
4.3: Flow model setup
An open channel flow model of the Tuul River is one-dimensional, distributed2, mathematical3,
conceptual4 and non-steady state5 hydraulic model.
The following flow chart is shown general sequence of flow modelling. All input data can divided
into two different types:
− Field data measured in field
− Non-field data assumed, constant and adjustable values
2 Model domains are discretised in space by use of uniform or non-uniform grid elements. 3 Partial differential equations are used in model. 4 Mathematical relations are applied to simulate the observed real world behavior. 5 A time variable is calculated for each calculation time step.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 38
Figure 27: Flowchart of a flow model
The flow of the Tuul river system has been modelled, calibrated and validated using some existing
data between 2005 and 2006. The following datasets are used to simulate a flow model.
1. Geographical locations of hydraulic stations, WTP and true value of distance between
stations, discharge points
2. Daily discharge data at Tuul-Ulaanbaatar station in 2005-2006
3. Daily discharge data at Tuul-Altanbulag station in 2005-2006
4. Monthly lumped discharge data of CWTP
5. Monthly lumped discharge data of Bio WTP
6. Hydraulic data at Tuul-Ulaanbaatar station
7. Hydraulic data at Tuul-Altanbulag station
8. Simplified cross-section of the Tuul River under Zaisan bridge
9. Simplified cross-section of the Tuul River under Altanbulag bridge
4.3.1: Network development
Network of the Tuul River model consists of the following objects:
− Two pink nodes are representing begin and end boundaries
− Sections
− Cross-sections (simplified)
− Two discharge points are indicating CWTP and Bio WTP
− Most lower node is demonstrating Tuul-Altanbulag station, that was used for calibration
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 39
Figure 28: A flow model network
A flow model network contains three layers; green areas are representing settlement areas, blue
line is the Tuul River and red squire is located in real position of CWTP. A begin node, right side,
is situated in 643389 E, 5305700 N and an end node, left side, is placed in 587593 E, 5301654 N.
Position of nodes (except for an end node), discharge points, length of sections (sum of length is
approximately 54 km), are all true values. However, position of an end node and length of last
section are false value that is assumption. A flow model setup is shown in Appendix 7.
Figure 29: Simplified cross-section of the Tuul River under Zaisan Bridge
Type of simplified upper cross section is trapezoid and floor level is 1285.25 m, surface level is
1292.6 m. Floor width of river channel is 48.6 m, maximum depth is 1.5 m and approximate area is
95.4 m2.
This simplified cross-section based on hydraulic measurement at Tuul-Ulaanbaatar station, which
implemented by WS, NAMHEM. Minimum and maximum depth, width of river is used to develop
simple cross-sections. For example, minimum depth of river was 0.38 m and maximum was 0.98 m.
Width of river at that time was 55.9 m and 67.7 m, correspondingly. The way of developing trapezoid
cross-section is placing minimum and maximum depth, width and draw straight lines to connect two
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 40
depths. Finally, slope angle calculated using a linear equation 15 and used this simplified cross-
section in DMS model. In this case, slope angle is 5.710.
Equation 15: Slope angle equation
−−
=∠12
12
xx
yyarctanS
Figure 30: Simplified cross-section of the Tuul River under Altanbulag Bridge
Type of middle cross section is trapezoid and floor level is 1178.3 m, surface level is 1182 m.
Floor width is 26.6 m, maximum depth is 1.5 m, slope angle is 5.40 and approximate area is
63.9 m2. Cross-sectional data interpolated along the river and slope of riverbed is 0.00198.
Note: If study area located higher than 1000 meter above sea level, then this software cannot simulate it.
This is one weakness (bug) of Duflow modelling system, which we discovered during our research.
This limitation not reported in the user guide of DMS. Then we decided to deduct 1000 from level
value and used it into model. For example, our actual value of floor level is 1285.25 m, but we
used 285.25 m instead of real one. Actually, this change has no effect in model. Reason of that is
difference of level values is used to calculate slope of riverbed.
To turn a network schematization in a flow model the following three steps has to be taken:
− Define initial condition
− Define boundary condition
− Configure the calculation [DMS, 2004b]
4.3.2: Initial condition of flow model
To start the computations, initial values for all state variables are required. These initial values
supplied for each node [DMS, 2004a]. Values, which used in initial condition, obtained from field
measured data and former computations.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 41
Table 17: Initial condition in flow model
Model structure Sampling point Discharge in m3 s-1 Level in m
Section 1 – begin Tuul-Zaisan 0.198 286.44
Section 1 – end Tuul-Sonsgolon 0.172 265.54
Section 2 – begin Tuul-Sonsgolon 0.172 265.54
Section 2 – end Tuul-Songino (upper) 0.155 249.7
Section 3 – begin Tuul-Songino (upper) 0.155 249.7
Discharge point 1 – begin CWTP 0.155 249.7
Discharge point 1 – end CWTP 0.155 249.7
Section 3 – end Tuul-Songino (down) 2.039 247.82
Section 4 – begin Tuul-Songino (down) 2.039 247.82
Discharge point 2 – begin Bio WTP 2.036 245.83
Discharge point 2 – end Bio WTP 2.036 245.83
Section 4 – end Tuul-Chicken farm 2.025 228.02
Section 5 – begin Tuul-Chicken farm 2.025 228.02
Section 5 – end Tuul-Khadanhyasaa 2 212.18
Section 6 – begin Tuul-Khadanhyasaa 2 212.18
Section 6 – end Tuul-Altanbulag 1.863 184.17
Section 7 – begin Tuul-Altanbulag 1.863 184.17
Section 7 – end Extension part 1.38 174.8
4.3.3: Boundary condition of flow model
In the flow model, boundary conditions should be defined at beginning and end nodes. Discharge
of the Tuul River at Tuul-Ulaanbaatar station, which measured in 2005 and 2006 by WS,
NAMHEM, is a boundary condition of beginning node. The type of boundary condition is an
equidistant (daily) starts from 01 April 2005 until 30 November 2006. Input data covers only warm
session. Reason of that is river freezes during cold session and consequence of frozen river is no or
less than 0.1 m3 s-1 discharge.
Figure 31: Discharge of the Tuul River under Zaisan Bridge in 2005 and 2006
0.000
20.000
40.000
60.000
80.000
100.000
120.000
140.000
160.000
3/1
/05
4/1
/05
5/1
/05
6/1
/05
7/1
/05
8/1
/05
9/1
/05
10/1
/05
11/1
/05
12/1
/05
1/1
/06
2/1
/06
3/1
/06
4/1
/06
5/1
/06
6/1
/06
7/1
/06
8/1
/06
9/1
/06
10/1
/06
11/1
/06
12/1
/06
Date
Q in
m3/s
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 42
An artificial water discharge outflow was selected as a boundary condition of the end node and
type is constant in 10 days. Time step and covered period are same with previous.
Figure 32: Artificial discharge at end node
-35.00
-30.00
-25.00
-20.00
-15.00
-10.00
-5.00
0.00
3/1
/20
05
4/1
/20
05
5/1
/20
05
6/1
/20
05
7/1
/20
05
8/1
/20
05
9/1
/20
05
10/
1/2
005
11/
1/2
005
12/
1/2
005
1/1
/20
06
2/1
/20
06
3/1
/20
06
4/1
/20
06
5/1
/20
06
6/1
/20
06
7/1
/20
06
8/1
/20
06
9/1
/20
06
10/
1/2
006
11/
1/2
006
12/
1/2
006
Date
Q in
m3/
s
4.3.4: Calculation setting
Before running model, calculation setting should be set and computation starts same with input
data. In this prototype model, the following calculation setting was defined:
− Starting from 01 March, 2005 till 1 December, 2006
− Time step is 1 minute and output is daily
− Theta equals to 0.9
− Distance between calculation points is 200 m
− Advection term is “Total”
− Resistance formula is Manning
Data in March was used to stabilize the model calculation. Output of model starts from 1st of April,
2005.
4.4: Calibration
Actual discharge data at Tuul-Altanbulag station were used to calibrate the flow model. Discharge
data, which used in calibration, measured by WS, NAMHEM between 1st of April and 30th of
November 2005. Rest of months are belongs to cold session. Hydraulic measurements do not take
place in the cold session.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 43
Figure 33: Correlation between calculated and observed discharge in calibration year
R2 = 0.8268
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00
Observed Q in m3/s
Cal
cula
ted
Q in
m3/
s
For the calibration year, correlation between calculated and observed values was evaluated by
different statistical approaches and the following results calculated in 2005 at Tuul-Altanbulag
station:
− ME -1.246
− MAE 2.539
− RMSE 4.162
− R2 0.827
− E 0.763
Figure 34: Observed and modelled Q in calibration year
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005
Date
Q in
m3/s
Observed Modelled
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 44
4.4.1: Sensitivity analysis
The main purpose of a sensitivity analysis is to quantify how sensitive the model is to certain
changes in the input data. A sensitivity analysis basically is a procedure to quantify on the
uncertainty of the calibrated model [Rientjes, 2006].
Sensitivity analyse of the flow model was done by using different Manning values. Results of
analysis evaluated in dissimilar statistical methods such as MAE, RMSE, R2, Nash-Sutcliffe model
efficiency coefficient and ME.
Figure 35: Sensitivity analysis
0.000
10.000
20.000
30.000
40.000
50.000
60.000
70.000
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005
Date
Q in
m3/s
Observed Modelled, n=0.04 Modelled, n=0.15
Manning values are less sensitive. However, some small changes were presented during analysis.
Figure 35 shown graphical representation of comparison observed and modelled values with
selected minimum and maximum Manning values.
Table 18: Statistical evaluation of sensitivity analysis
Different Manning values Statistical evaluation n=0.03 n=0.04 n=0.06 n=0.1 n=0.15 R2 n.a 0.821 0.828 0.837 0.841 ME n.a -1.218 -1.218 -1.218 -1.222 MAE n.a 2.534 2.520 2.501 2.503 RMSE n.a 4.192 4.128 4.040 4.035 E n.a 0.760 0.767 0.777 0.778
Totally, five different Manning values were selected for sensitivity analysis and evaluated by
different statistical methods. Based on graph and statistical assessment, the best result given by
n=0.15. Because of, R2 value is highest and RMSE is lowest one. Therefore, coefficient of model
efficiency is 0.778.
However, n=0.06 was applied in a flow model. Reason of that, characteristics of river channel is
coincide with 1f in table for the Manning values (see Appendix 8).
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 45
4.5: Validation
Actual discharge data at Altanbulag station was used to validate flow model. Discharge data, which
used in validation, measured by WS, NAMHEM between 1st of April and 30th of November 2006.
Figure 36: Correlation between calculated and observed Q data in validation year
R2 = 0.885
0.00
10.00
20.00
30.00
40.00
50.00
0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00
Observed Q in m3/s
Cal
cula
ted
Q in
m3/
s
In validation year, correlation between calculated and observed values was evaluated by different
statistical approaches and the following results calculated in 2006 at Tuul-Altanbulag station:
− ME 1.347
− MAE 2.677
− RMSE 4.781
− R2 0.885
− E 0.830
Figure 37: Observed and modelled Q in validation year
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
4/1/2006 5/1/2006 6/1/2006 7/1/2006 8/1/2006 9/1/2006 10/1/2006 11/1/2006
Date
Q in
m3
/s
Observed Modelled
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 46
4.6: Chapter conclusion
From a flow model simulation, the following conclusions can be drawn:
1. DMS is suitable for simulating the selected section of the Tuul River in Mongolia. Because of
the average slope of the riverbed is less than 1% (0.2%), yielding mostly sub-critical flow
regimes (i.e. Fr = 0.74), water levels and flows could be simulated with good accuracy in the
river reaches upstream and downstream of the Ulaanbaatar city.
2. Sensitivity analysis showed that the flow model is not very sensitivity to Manning values.
Difference between selected minimum and maximum values is R2=0.02 and E=0.018.
3. If real water level and river bed elevation values are greater than one thousand (m.a.s.l.), then
DMS shows an error message. Maybe, this is a bug of this software. If observed real data
values are used, they have to be rescaled to (<1000 m) values. It has minor effect to the
quality model in terms of re-aeration.
4.7: Limitation in a flow modelling
In the flow model simulation, interpretation of actual water level data in both stations, especially at
Tuul-Altanbulag station, was sometimes very confusing. According to some sources, artificial surface
level were used to measure the water levels. For example, we received values greater than 5 meter,
typed in water level data at Tuul-Altanbulag station. In reality and proven by field inspection, those
values were false. Water level measurements and artificial surface level data could therefore not be
used in the flow model calibration and validation experiment.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 47
Chapter 5: Water quality model
5.1: Introduction of quality model
The DMS water quality model is usable for simulating the transportation of substances in free surface
quality. One can define its own substances and determine it is in or decline in concentration. There is
a standard parameter for natural, algorithmic decay depending on actual concentration. In DMS water
quality a distinction is made between substances, which are transported with the quality of the water,
for instance dissolved substances, and bottom materials that are not transported. This distinction
offers the user the facility to, for instance, study the interaction between the bottom materials and the
dissolved substances in the water above [DMS, 2007].
DMS is a flexible package, as the user is able to define his own set of process descriptions. This
means that except for the transport of pollutants within the network, the user can provide all other
physical, chemical and biological interactions between the state variables. The input for the
simulation of the water quality depends on the defined state variables and processes. Typically,
parameters like process rates, equilibrium constants and other model constants have to be entered.
Finally, external variables can be entered as constraints or as time functions (i.e. temperature and
irradiance) [Makkinga, et al., 1998].
Water quality modelling can be a valuable tool for water management since it can simulate the
potential response of the aquatic system to such changes as the addition of organic pollution, the
building of small hydro-electric power plants, the increase in nutrient levels or water abstraction rates
and changes in sewage treatment operations (such as the addition of tertiary treatment). Most existing
river models are for oxygen balance and are based on BOD measurements. Krenkel and Novotny
(1980) have listed the categories of variables required for oxygen balance modelling as:
− hydrological variables (e.g. river discharge),
− hydraulic variables (velocity, geometry of river bed, turbulence, etc.),
− oxygen sinks (e.g. benthic oxygen demand, nitrification of ammonia),
− oxygen sources (e.g. re-aeration, atmospheric exchange, primary production), and
− temperature
Downstream of sewage effluent discharges from treatment plants using biological, secondary
processes, bacterial activity may also need to be incorporated into the models [Krenkel and Novothy,
1980] and [WHO, 1996].
In system analysis, the processes affecting the oxygen concentration at a certain time at an exact point
in a water body can be schematized as:
− advection and dispersive mechanisms of oxygen transport in the water body
− exchange with the atmosphere (re-aeration)
− oxygen consumption for oxidation of organic matter by micro-organisms;
− biochemical oxygen demand and nitrification
− Oxygen demand of bottom sediment and benthic micro-organisms, SOD
− Primary production and respiration of photosynthetic i.e., the oxygen production and demand
of algae and water plants [Mannaerts, 2007]
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 48
Concentrations of dissolved oxygen in unpolluted waters range usually between 7 mg l-1 in warmer
and 14 mg l-1 in colder waters as they are a strong function of water temperature. Variations in DO
can occur seasonally, or even over 24 hour periods, in relation to temperature and biological activity
(i.e. photosynthesis and respiration). Biological respiration, including that related to decomposition
processes, reduces DO concentrations. In still waters, pockets of high and low concentrations of
dissolved oxygen can occur depending on the rates of biological processes. Waste discharges high in
organic matter and nutrients can lead to decreases in DO concentrations as a result of the increased
microbial activity (respiration) occurring during the degradation of the organic matter [WHO, 1996].
Sources of DO:
− Re-aeration from the atmosphere
− Photosynthetic oxygen production
− DO in tributaries or effluents
Sinks of DO:
− Oxidation of carbonaceous waste material
− Oxidation of nitrogenous waste material
− Oxygen demanding sediments of water body
− Use of oxygen for respiration by water plants [Thomann and Mueller, 1987]
5.2: Background theory of quality modelling
The mass transport equation, Peclet number, formulas for the transport of pollutants within the
network and physical, chemical, biological interactions between the state variables are part of the
mathematical background of quality modelling.
5.2.1: Mass transport equation
The quality part of the DUFLOW package is based upon the one dimensional transport equation.
This partial differential equation describes the concentration of a constituent in a one-dimensional
system as function of time and place.
Equation 16: Momentum balance equation
( )P
x
CAD
xx
)QC(
t
BC +
∂∂
∂∂+
∂∂−=
∂∂
Where:
C constituent concentration [g/m3]
Q quality [m3/s]
A cross-sectional quality area [m2]
D dispersion coefficient [m2/s]
B cross-sectional storage area [m2]
x x coordinate [m]
t time [s]
P production of the constituent per unit length of the section [g/m.s]
The production term of the equation includes all physical, chemical and biological processes to
which a specific constituent is subject. In order to apply this method equation 16 is rewritten as:
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 49
Equation 17: Mass balance equation
( )0P
t
BC
x
S =−∂
∂+∂∂
In which, S is the transport (quality of the constituent passing a cross-section per unit of time):
Equation 18: Equation for the constituent transport by advection and dispersion
x
CADQCS
∂∂−=
Equation 18 describes the transport by advection and dispersion. Equation 17 is the mathematical
formulation of the mass conservation law, which states that the accumulation at a certain location x
is equal to the net production rate minus the transport gradient [DMS, 2004a].
The dispersion coefficient either can be supplied by the user or can be calculated from the
characteristics of the quality. The empirical equation according to Fisher (1979) is used to
calculate the quality dependent part. In order to prevent the dispersion coefficient to become 0 at
low quality velocities a constant term is added that reflects background dispersion. The quality
independent part is in particular important in stagnant systems, where it represents the wind
induced mixing. The following equation is used:
Equation 19: Background dispersion coefficient
( ) 0
22s
x Du*Z
W*ut,xD +α=
Where:
αx a proportionality constant [-]
W quality width [m]
us average quality over the cross-sectional area [m/s]
Z water depth [m]
u shear stress velocity [m/s]
D0 background dispersion coefficient [m2/s]
The shear stress u can be written as:
Equation 20: Shear stress equation
C
guu s=
With:
C coefficient of De Chézy [m1/2/s]
g acceleration due to gravity [m2/s]
As all the characteristics of the quality are known or calculated in the quality part, above-
mentioned equations easily can be used to calculate the dispersion coefficient. Only αx and D0
have to be supplied by the user [DMS, 2004a].
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 50
5.2.2: Peclet number
In fluid dynamics, Peclet number is a dimensionless number relating the rate of advection of a flow
to its rate of diffusion. It is equivalent to the product of the Reynolds number with the Schmidt
number in the case of mass diffusion. For mass diffusion, it is defined as:
Equation 21: Peclet number
D
LVPe=
Where;
L characteristic length
V velocity
D mass diffusivity [Patankar, 1980]
5.2.3: DO balance equations
The amount of oxygen in the water is function of re-aeration, photosynthesis process, sediment
oxygen demand, oxygen consumption for biochemical process and nitrification.
Figure 38: Schematization of DO balance model
The following equations are used to DO model and initially developed by University of
Wageningen.
The DO balance in the river can be written as follows:
Equation 22: Dissolved oxygen balance
( ) ( ) ( ) ( ) ( ) ( )dt
Nitrifd
dt
BODoxd
dt
SedDOd
dt
PDOd
dt
OSd*
dt
KAd
dt
dDO ++++=
Where:
DO dissolved oxygen concentration in the water in mg l-1
KA re-aeration rate in day
OS oxygen saturation in mg l-1
PDO primary production in g m-2 day--1
SedDO oxygen consumption for sediment in g m-2 day-1
BODox oxidation of BOD in g m-2 day-1
Nitrif oxygen consumption for nitrification in g m-2 day-1
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 51
Re-aeration defined as:
− Process of oxygen absorption from the atmosphere
− Assumed a first order process; absorption rate is proportional to deficit
A basic and main theory of gas transfer, which commonly used in flowing waters such as river, is a
surface renewal model. Danckwerts (1951) modified Higbie’s (1935) the penetration theory by
assuming that the fluid elements reach and leave the interface randomly. The direction and
magnitude of the mass transfer depends partially on the difference between the saturation value
and the actual value in the water. If the water is unsaturated (o < os), then transfer will be positive
(a gain) as oxygen moves from the atmosphere into the water to try to bring the water back to the
equilibrium state of saturation. Conversely, if the water is supersaturated (o > os), then transfer
will be negative (a loss) as oxygen is purged from the system [Chapra, 1997].
The following series equations used for re-aeration rate:
Equation 23: Temperature dependent oxygen mass-transfer velocity
)20T(TKL*dt
20KL
dt
KLT −=
Equation 24: Re-aeration rate
Z
KLT
dt
KA =
Where:
TKL temperature coefficient of oxygen mass-transfer (1.024)
KLmin minimum oxygen mass-transfer rate in flowing system in m d-1 (0.1)
KL20 oxygen mass-transfer rate in the water laminar layer in m d-1
KLT temperature dependent O2 mass-transfer velocity in the water laminar layer in m d-1
KA re-aeration rate in day
When:
If KL20 < KLmin, then KL20 = KLmin
Oxygen saturation can be solved by: The saturation concentration of oxygen in natural water is on the order of 10 mg l-1. In general,
several environmental factors can affect this value. From the perspective of water quality
modelling, the most important of these are:
− Water temperature
− Salinity
− Partial pressure variations due to elevation
Several empirical derived equations have been developed to predict how these factors influence
saturation [Chapra, 1997]. However, only water temperature taken into account and the following
equation can be used to calculate the dependence of oxygen saturation on water temperature (0C):
Equation 25: Oxygen saturation
( ) ( ) ( ) ( ) ( ) ( ) ( )dt
Td*
dt
Td*
dt
Td*000077774.0
dt
Td*
dt
Td*007991.0
dt
Td*41022.0652.14
dt
OSd −+−=
With:
T water temperature in 0C
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 52
Primary production written as below: Photosynthesis creates oxygen and respiration depletes oxygen, the plants will have an impact on
the oxygen resources. Because photosynthesis is light dependent, this effect can have both
seasonal and diurnal manifestations. In most water quality models, the rate of photosynthesis is
assumed to be directly proportional to light energy. Oxygen level could be supersaturated during
the afternoon and severely depleted just before dawn [Chapra, 1997]. Primary production by
photosynthetic systems is functions of solar radiation, algal biomass and can be presented by:
Equation 26: Primary production
( )A*IO*
dt
PDOd β=
Where:
β oxygen production constant in (gO2 day-1. mg Chl-1) (W.m-2)-1
IO light intensity in W m-2
A chlorophyll concentration in ug l-1
Oxygen consumption for sediment defined as below; SOD usually expressed as a distributed sink term in g m-2 day-1. Sediment oxygen demand is due to
the oxidation of organic matter in bottom sediments. These benthic deposits or sludge beds derive
from several sources. Wastewater particulates, leaf litter and eroded organic rich soil can result in
sediments with high organic content. Regardless of the source, oxidation of the accumulated
organic matter will result in a SOD. Zison et al, (1978) have reported a range of 1.04 to 1.13 for
TSOD. A value of 1.065 is commonly employed [Chapra, 1997].
Equation 27: Oxygen consumption for sediment
( ) ( )
Z
TSOD*SOD
dt
SedDOd 20T−−=
With;
SOD sediment oxygen demand in g m-2 day-1
TSOD temperature coefficient of SOD (1.06)
Oxidation of BOD written as: The analysts introduced sewage sample into a bottle and merely how much oxygen was consumed.
The resulting quantity was dubbed BOD. Water quality analysts early on adopted 5-day BOD test.
Incubation time 5 days makes the test practical, then to extrapolate the 5-day result to the ultimate
BOD level. This is usually done by performing a long term BOD to estimate the decay rate
[Chapra, 1997].
Equation 28: Ultimate BOD
5*Kdexp1
BOD
dt
BODU−−
=
BOD represents the oxygen demand equivalent to the complex biodegradable organic matter
present in water. First order reaction kinetics is used to decay the matter [Mannaerts, 2007].
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 53
Equation 29: Biochemical oxygen demand
( ) ( )KNDODO
DO*TKd*BODU*Kd
dt
BODoxd 20T
+−= −
Where;
Kd BOD degradation rate constant in day
BODU ultimate BOD in mg l-1
TKd temperature coefficient of BOD decay (1.05)
KDO Monod constant DO inhibition BOD decay in mg l-1
Nitrification rate solved by the following: Nitrogen compounds in water also have an impact on oxygen resources. Sewage nitrogen can be
broadly broken down into organic nitrogen compounds and ammonia autotrophic bacteria then
assimilate the ammonia and create nitrite and nitrate. The oxidation would consume 4.57 g of
oxygen per 1 g of Kjeldahl nitrogen [Chapra, 1997]. In terms of oxygen required to complete the
nitrification steps form ammonia to nitrate, the equation can be written as:
Equation 30: Oxygen consumption for nitrification
( ) ( )DOKNDO
DO*TKnit*4NH*Knit*57.4
dt
Nitrifd 20T
+−= −
With;
Knit nitrification rate constant in day
NH4 ammonium in mg l-1
TKnit temperature coefficient of nitrification (1.05)
KNDO Monod constant DO inhibition nitrification in mgl-1 [Lijklema, et al., 1996]
5.3: Quality model setup
This water quality model of the Tuul River is a one-dimensional, distributed, mathematical,
conceptual and non-steady state, prototype model. The quality model is linked to the water quantity or
flow model, mentioned in a previous chapter. Therefore, a model network, boundary and initial
conditions are reorganized using output of a flow model. The DO balance of the Tuul River system
has been modelled, calibrated and validated using some existing hydro-chemical data between 2005
and 2006. The following hydro-chemical field datasets were added into a flow model.
1. Monthly data of DO, BOD5, NH4+ at Tuul-Zaisan sampling point
2. Monthly data of DO, BOD5, NH4+ at Tuul-Sonsgolon sampling point
3. Monthly data of DO, BOD5, NH4+ at Tuul-Songino (upper) sampling point
4. Monthly data of DO, BOD5, NH4+ at Tuul-Chicken farm sampling point
5. Monthly data of DO, BOD5, NH4+ at Tuul-Khadanhyasaa sampling point
6. Monthly data of DO, BOD5, NH4+ at Tuul-Altanbulag sampling point
7. Monthly average data of DO, BOD5, NH4+ from the CWTP
8. Hourly average solar radiation data at Ulaanbaatar station downloaded from global open
source
9. Daily average wind speed data from open source
10. Daily average water temperature data at Tuul-Ulaanbaatar station
Rest of data belongs to non-field.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 54
A flowchart has shown general structure and procedure of a quality model.
Figure 39: Flowchart of a quality model
Figure 40: Quality model network schematization
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 55
DO modelling network of the Tuul River model consists of the following objects:
− Nodes are representing sampling points along the river
− First 6 sections are demonstrating the Tuul River
− Cross-sections are interpolated along the river
− Two discharge points are indicating CWTP and Bio WTP.
Position of nodes, discharge points and length of the river are all-true. On the other hand, position of
an end node and length of last section are false values that used in a flow model.
To make a quality model out of a flow model, the following steps need to be carried out:
− Define quality description file
− Define initial conditions for quality
− Define boundary conditions for quality
− Define parameters
− Define external variables
− Configure the calculation [DMS, 2004b]
5.3.1: Quality description
A quality description file consists of main two parts:
− Declaration part:
The different variables are defined.
− Compound statement for water courses:
The equations are describing the processes in the water route.
Four types of variables are distinguished in quality description.
1. water quality has effect on this variables.
2. xt external variables, which are space and time dependent.
3. parm parameters, constants and coefficients used in process.
4. flow supplied by the hydraulic part of model [DMS, 2004a].
The following chemical and hydraulic variables were used in DO model:
− WATER DO Dissolved oxygen
− WATER BOD Biochemical oxygen demand
− WATER NH4 Ammonium
− FLOW Q Discharge
− FLOW As Cross-section of flow area
− FLOW Z Water depth
The quality description file, which utilized in this quality model, initially developed in Department
of Water Quality Management and Aquatic Ecology, Agricultural University of Wageningen. See
Appendix 9.
5.3.2: Initial condition of quality model
For all objects, initial concentrations must be defined. Initial values in discharge and level columns
obtained from flow model. Besides, values in chemical columns are collected from field data,
which measured by CLEM.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 56
Table 19: Initial condition in quality model
Model structure Sampling point BOD in mg l-1 DO in mg l-1 NH4+ in mg l-1
Section 1 – begin Tuul-Zaisan 6 9.94 0.7 Section 1 – end Tuul-Sonsgolon 6.1 8.48 0.78 Section 2 – begin Tuul-Sonsgolon 6.1 8.48 0.78 Section 2 – end Tuul-Songino (upper) 6 7.42 1.51 Section 3 – begin Tuul-Songino (upper) 6 7.42 1.51 Discharge point 1 – begin CWTP 30.6 4.41 29.33 Discharge point 1 – end CWTP 30.6 4.41 29.33 Section 3 – end Tuul-Songino (down) 62.8 0.9 1.4 Section 4 – begin Tuul-Songino (down) 62.8 0.9 1.4 Discharge point 2 – begin Bio WTP 62.8 0.9 1.4 Discharge point 2 – end Bio WTP 62.8 0.9 1.4 Section 4 – end Tuul-Chicken farm 26.9 1.14 1.94 Section 5 – begin Tuul-Chicken farm 26.9 1.14 1.94 Section 5 – end Tuul-Khadanhyasaa 9.6 4.4 1.58 Section 6 – begin Tuul-Khadanhyasaa 9.6 4.4 1.58 Section 6 – end Tuul-Altanbulag 6 7.5 1.05 Section 7 – begin Tuul-Altanbulag 6 7.5 1.05 Section 7 – end Extension part 6 7.5 1.05
5.3.3: Boundary condition of quality model
Quality boundary conditions can be entered as time series for concentration. In quality model,
boundary conditions should be defined at nodes and discharge points that have flow boundary
condition.
Table 20: Quality boundary condition
Sampling points
Var
iab
les
in m
g/l
Mar
ch-0
5
Ap
ril-
05
May
-05
Jun
e-0
5
July
-05
Au
gu
st-0
5
Sep
tem
ber
-05
Oct
ob
er-0
5
No
vem
ber
-05
Dec
emb
er-0
5
Jan
uar
y-0
6
Feb
ruar
y-0
6
Mar
ch-0
6
Ap
ril-
06
May
-06
Jun
e-0
6
July
-06
Au
gu
st-0
6
Sep
tem
ber
-06
Oct
ob
er-0
6
No
vem
ber
-06
DO 10 12 6 9 8 9 10 12 n.a n.a n.a n.a n.a 8 8 8 8 9 9 13 13 BOD 6 3 1 3 2 0 1 4 n.a n.a n.a n.a n.a 1 2 1 1 1 1 4 2 Tuul-Zaisan NH4
+ 1 0 0 0 0 0 0 0 n.a n.a n.a n.a n.a 0 0 0 0 0 0 0 0 DO 8 11 8 9 8 11 9 12 11 5 n.a n.a n.a n.a 9 9 9 9 12 11 n.a BOD 6 1 1 3 1 1 1 4 2 2 n.a n.a n.a n.a 2 3 1 1 1 4 n.a
Tuul-Sonsgolon
NH4+ 1 0 0 0 0 0 0 0 0 0 n.a n.a n.a n.a 0 0 0 0 0 0 n.a
DO 7 9 8 9 9 11 11 12 11 n.a n.a n.a n.a 8 9 9 8 9 9 10 11 BOD 6 2 2 2 1 1 2 4 2 n.a n.a n.a n.a 2 2 3 0 1 2 3 1
Tuul-Songino (upper)
NH4+ 2 0 0 0 0 0 1 0 0 n.a n.a n.a n.a 0 0 0 0 0 0 0 0
DO 1 2 9 10 8 9 11 12 6 n.a 4 n.a n.a 3 8 8 6 8 8 8 10 BOD 27 34 2 7 7 5 4 5 35 n.a 21 n.a n.a 52 2 6 2 6 5 14 21
Tuul-Chicken farm
NH4+ 2 12 1 1 1 1 1 2 13 n.a 31 n.a n.a 14 0 1 1 0 1 3 10
DO 4 6 11 9 8 9 11 12 8 3 5 n.a 6 5 8 8 6 8 8 10 8 BOD 10 17 6 8 6 5 3 5 21 9 13 n.a 18 18 4 6 3 5 6 6 4
Tuul-Khadanhyasaa
NH4+ 2 8 1 1 1 1 0 2 6 9 13 n.a 14 9 1 1 0 0 1 0 4
DO 8 10 10 9 8 9 12 12 9 4 n.a n.a 8 4 8 9 6 9 8 11 10 BOD 6 8 4 8 6 3 3 6 3 4 n.a n.a 5 14 6 8 1 5 6 10 2
Tuul-Altanbulag
NH4+ 1 5 1 0 0 1 0 1 5 5 n.a n.a 7 8 1 0 0 0 1 0 4
DO 4 2 4 2 2 2 1 n.a 7 n.a 2 2 1 1 0 2 4 4 3 3 3 BOD 31 24 18 33 31 29 34 32 36 43 49 29 38 52 60 19 20 21 19 27 41 CWTP NH4
+ 29 19 17 16 22 21 21 17 19 16 17 20 20 25 28 35 23 17 30 27 36
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 57
Above-mentioned hydro-chemical datasets, which were measured by CLEM and CWTP
respectively, are boundary conditions of nodes and discharge points. Type of boundary condition is
non-equidistant, starts from 01 April 2005 until 30 November 2006. Input data covers only warm
session. All values in boundary conditions are monthly and DMS automatically interpolated
between two values.
5.3.4: Parameters in quality model
In quality declaration section, parameters are abbreviated as “parm”. All constant numbers,
coefficients are belongs to the parameters. The following parameters are used in a prototype
quality model:
− PARM Beta Oxygen production constant (photosynthesis)
− PARM fd Fraction of dissolved BOD
− PARM Kd BOD degradation rate constant
− PARM KDO Monod constant DO inhibition BOD decay
− PARM KLmin Minimum oxygen mass-transfer rate in flowing system
− PARM KNDO Monod constant DO inhibition nitrification
− PARM Knit Nitrification rate constant
− PARM OPTKL Option material transfer
− PARM TKd Temperature coefficient of BOD decay
− PARM TKL Temperature coefficient of oxygen mass-transfer
− PARM TKnit Temperature coefficient of nitrification
− PARM TSOD Temperature coefficient of sediment oxygen demand
− PARM Vs Settling velocity of BOD
5.3.5: External variables
External variables can be defined in the quality model description for input data that are space and,
or time dependent [DMS, 2004b]. It is shortened to “xt”. External variables, which are used in this
model as follow:
− XT T Water temperature
− XT SOD Sediment oxygen demand
− XT SNH4 Diffusion of NH4 concentration
− XT SBOD Diffusion of BOD concentration
− XT I0 Solar radiation
− XT A Algal biomass
− XT W Wind speed
Besides of this, dispersion coefficient has to be described in the quality model itself. The value of
dispersion coefficient can be defined by the following an empirical equation [Chapra, 1997]:
Equation 31: Dispersion coefficient
BS
Q05937.0E
0x =
Where:
Ex dispersion coefficient in m2 s-1
Q average flow in m3 s-1
S0 riverbed slope
B river width in m
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 58
Above-mentioned equation proposed by McQuivey and Keefer in 1974. They proposed an
equation the following conditions where the Froude number is less than 0.5, flow ranging from 1
to 934 m3 s-1 and slope from 0.000045 to 0.00298 [Chapra, 1997].
In case of selected section of the Tuul River, average Fr is 0.34, discharge 22.64 m3 s-1 and slope is
0.001981.
Moreover, monthly mean water temperature, was implemented by CLEM and hourly average solar
radiation, monthly mean wind speed, which were measured by Ulaanbaatar weather station,
datasets are obtained from online open course IWEC and applied to this model [ASHRAE, 2001].
5.3.6: Calculation setting
Before running quality model, additional calculation setting should be set. Quality calculation
setting was defined as below:
− Time step is 30 minutes and output is daily
− Type of calculation is “Flow and Quality”
− Theta equals to 0.55
− Output variable is DO
5.4: Calibration
Monthly DO analysis data at sampling point Tuul-Songino (down) was applied to calibrate the quality
model. Chemical data, which used in calibration, measured by CLEM between April and December
2005. The motivation of select this sampling point to calibration, is that the station is the first
sampling point, after the CWTP effluent discharges into the Tuul River. See quality model network.
Figure 41: Correlation between calculated and observed DO in calibration year
R2 = 0.7936
0.00
2.00
4.00
6.00
8.00
10.00
12.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00
Observed Q in m3/s
Cal
cula
ted
Q in
m3
/s
In calibration year, correlation between calculated and observed values was evaluated by different
statistical approaches and the following results calculated in 2005 at Tuul-Songino (down) sampling
point:
− ME -0.76
− MAE 0.90
− RMSE 1.68
− R2 0.79
− E 0.72
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 59
Figure 42: Observed and modelled DO in calibration year
0.00
2.00
4.00
6.00
8.00
10.00
12.00
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005
Date
Q in
m3
/s
Observed Modelled
5.4.1: Sensitivity analysis
Sensitivity analyse of the quality model was completed by varying parameters such as β, fd, Kd,
KNDO, Knit, Vs, TKd, TKnit and SOD. In the analysis, minimum and maximum values of
selected parameters were used. Reason of select those parameters, is that the parameters values
have some range in the literature. If model is sensitive to selected parameters then such small
different values give dissimilar statistical results. From those analyses, some selected results are
represented in here. Results of sensitivity analysis were assessed using several statistical methods.
5.4.1.1: Very sensitive parameters
Beta is representing the oxygen production constant by photosynthesis process in the water, due
to algae (and or aquatic plants theoretically). Result of photosynthesis process, certain amount
of oxygen produced by algal biomass. BOD degradation rate constant and nitrification rate
constant are abbreviated as Kd and Knit, respectively.
Figure 43: Sensitivity analysis for beta
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005
Date
DO
in m
g/l Observed
B=0.0001
B=0.0005
B=0.001
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 60
Table 21: Statistical evaluation of very sensitive parameters
Statistical methods Parameter
values ME MAE RMSE E R2
β=0.0001 -0.76 0.90 1.68 0.72 0.79
β=0.0005 -31.69 31.69 36.01 -130.04 0.25
β=0.001 -70.52 70.52 80.72 -657.48 0.20
Kd=0.1 0.76 0.90 1.68 0.72 0.79
Kd=0.3 -2.43 2.47 2.98 0.11 0.71
Knit=0.1 -2.03 2.03 2.55 0.35 0.79
Knit=0.5 -1.04 1.04 1.90 0.64 0.76
Knit=1.0 -0.76 0.90 1.68 0.72 0.79
Based on above-mentioned graph and table, the model is very sensitive to β, Kd and Knit.
Because of sensitivity, changes have strong effect. The relations between parameters β, Kd and
DO are direct. This means, parameter value increases then DO value increases as well. Between
Knit and DO have inverse relation.
5.4.1.2: Less sensitive parameters
Temperature coefficients of BOD decay and nitrification are shortened as TKd and TKnit,
correspondingly. SOD and fd are short form of sediment oxygen demand and fraction of
dissolved BOD, respectively.
Figure 44: Sensitivity analysis for SOD
0.00
2.00
4.00
6.00
8.00
10.00
12.00
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005
Date
DO
in m
g/l
Observed
SOD=0.5
SOD=2.0
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 61
Table 22: Statistical evaluation of less sensitive parameters
Statistical methods Parameter values ME MAE RMSE E R2
TKd=1.03 -0.53 0.90 1.60 0.74 0.78 TKd=1.05 -0.76 0.90 1.68 0.72 0.79
TKnit=1.05 -0.76 0.90 1.68 0.72 0.79 TKnit=1.1 -0.98 1.06 1.94 0.62 0.74 SOD=0.5 -0.98 1.02 1.77 0.68 0.80 SOD=2.0 -0.76 0.90 1.68 0.72 0.79
fd=0.8 -0.84 0.92 1.70 0.71 0.80 fd=0.9 -0.80 0.91 1.69 0.71 0.80 fd=1.0 -0.76 0.90 1.68 0.72 0.79
The model is less sensitive to those parameters that shown in table and graph. Reason of that is
such changes have not strong effect. The relations between parameters TKd, TKnit and DO are
direct. SOD and fd parameters inversely relate to DO value. This means, SOD value increases
then DO concentration reduces.
5.4.1.3: Very less and non sensitive parameters
Monod constant DO inhibition nitrification and settling velocity of BOD are shortened as
KNDO and Vs, respectively.
Figure 45: Sensitivity analysis for Vs
0.00
2.00
4.00
6.00
8.00
10.00
12.00
4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005
Date
DO
in m
g/l
Observed
Vs=0.1
Vs=1.0
Table 23: Statistical evaluation of not sensitive parameters
Statistical methods Parameter
values ME MAE RMSE E R2
KNDO=1.0 -0.76 0.90 1.68 0.72 0.79
KNDO=1.5 -0.78 0.91 1.71 0.71 0.79
KNDO=2.0 -0.80 0.92 1.73 0.70 0.78
Vs=0.1 -0.76 0.90 1.68 0.72 0.79
Vs=1.0 -0.76 0.90 1.68 0.72 0.79
Based on results, KNDO is very less sensitive and the model is non-sensitive to Vs. Reason of,
that is Vs was used to calculate BOD concentration.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 62
5.5: Validation
Actual DO measurement, which was implemented by CLEM, applied to validate a prototype quality
model. Validation data covers between April and December 2006.
Figure 46: Correlation between calculated and observed Q data in validation year
R2 = 0.8779
0.00
2.00
4.00
6.00
8.00
10.00
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
Observed Q in m3/s
Cal
cula
ted
Q in
m3/s
In validation year, correlation between calculated and modelled values was evaluated by different
statistical approaches and the following results calculated in 2006 at Tuul-Songino (down) sampling
point:
− ME -0.01
− MAE 0.45
− RMSE 0.87
− R2 0.88
− E 0.87
Figure 47: Observed and modelled Q in validation year
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
4/1/2006 5/1/2006 6/1/2006 7/1/2006 8/1/2006 9/1/2006 10/1/2006 11/1/2006 12/1/2006
Date
Q in
m3
/s
Modelled Observed
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 63
5.5.1: Scenarios for SWQ improvement
The quality model has been used to evaluate the impact of scenarios for improvement of water
quality in order to meet the Mongolian National Standard. The national standard declares that,
permissible level of DO concentration must equal or above 6 mg l-1 in warm session [CSM, 1998].
A threshold value of 6 mg l-1 was selected in this scenario. Tuul-Songino (down) sampling point
(same as calibration and validation) was selected for checking point and is close to the discharge
point from the CWTP.
If assuming, river discharge and DO, BOD5, NH4+ concentration in the upstream reach, before
CWTP effluent pours into the Tuul River, to be the same as in 2005 at Tuul-Songino (upper)
sampling point, the following options can be suggested:
− Option 1 for ammonium: DO =< 4 mg l-1, BOD5 => 17 mg l-1, NH4+ => 8 mg l-1
− Option 2 for BOD: DO =< 4 mg l-1, BOD5 => 17 mg l-1, NH4+ => 15 mg l-1
− Option 3 for DO: DO =< 8 mg l-1, BOD5 => 17 mg l-1, NH4+ => 15 mg l-1
The concentration values in the options were taken from minimum and maximum concentration
values of CWTP discharge in 2005. That means, CWTP discharge contained NH4+
min=15.74 mg l-1,
DOmax=6.93 mg l-1, BODmin=17.89 mg l-1 in 2005. In other words, the CWTP has a “serious”
possibility to achieve this point.
Figure 48: Options for SWQ improvement
0
2
4
6
8
10
12
14
4/1/
05
5/1/
05
6/1/
05
7/1/
05
8/1/
05
9/1/
05
10/
1/0
5
11/
1/0
5
12/
1/0
5
Date
DO
in m
g/l
O1; O=4, B=17, N=8
O2; O=4, B=17, N=15
O3; O=8, B=17, N=15
Threshold value
Monthly average O1
Monthly average O2
Monthly average O3
Observed in 2005
Option 1:
If CWTP discharge contains DO =< 4 mg l-1, BOD => 17 mg l-1, and NH4+ => 8 mg l-1 then DO in
the Tuul River improves as below:
Table 24: Scenario for SWQ improvement, Option 1
Month April May June July August September October November Simulated 8.7 7.7 9.3 10.9 12.8 12.3 10.8 7.9 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +6.1 -0.7 0 +2.4 +3.0 +1.6 +0.6 +3.9
Relative difference, % -234.6 +8.3 0 -28.2 -30.6 -15.0 -5.9 -97.5 RMS 3.0
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 64
Option 2:
If CWTP discharge contains DO =< 4 mg l-1, BOD => 17 mg l-1, and NH4+ => 15 mg l-1 then DO
in the Tuul River improves as below:
Table 25: Scenario for SWQ improvement, Option 2
Month April May June July August September October November Simulated 8.5 6.9 8.8 10.2 11.9 11.5 9.9 6.6 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +5.9 -1.5 -0.5 +1.7 +2.1 +0.8 -0.3 +2.6
Relative difference, % -226.9 +17.9 +5.4 -20.0 -21.4 -7.5 +2.9 -65.0 RMS 2.6
Option 3:
If CWTP discharge contains DO =< 8 mg l-1, BOD => 17 mg l-1, and NH4+ => 15 mg l-1 then DO
in the Tuul River improves as below:
Table 26: Scenario for SWQ improvement, Option 3
Month April May June July August September October November Simulated 9.5 7.1 8.9 10.3 12.0 11.6 10.1 6.8 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +6.9 -1.3 -0.4 +1.8 +2.2 +0.9 -0.1 +2.8
Relative difference, % -265.4 +15.5 +4.3 -21.2 -22.4 -8.4 +1.0 -70 RMS 2.9
Based on these figure, tables, suggested values and statistical analysis, the best option is the first.
Because of, monthly mean DO concentrations are not lower than the threshold value and monthly
improvement are greater than other options. Those options are valid only for the warm session.
Obviously, discharge from CWTP should contain DO concentration greater and BOD5, NH4+
concentrations less than the selected option in cold session.
5.6: Chapter conclusion
Some conclusions drawn from the quality modelling are mentioned as below:
− Quality model syntax (oxygen model) can be used for this river.
− Very limited number of observations (once per month) has an effect on the results, such as
high statistical results.
− In sensitivity analysis, the model is very sensitive to β, Kd and Knit. Parameters TKd, TKnit,
SOD and fd are less sensitive. Parameter KNDO is very less and Vs is not affective to the
model. Relations between DO concentration and β, Kd, KNDO, TKd, TKnit are direct.
Inverse relations have between DO and fd, Knit, SOD.
− Best option for SWQ improvement is first (see before). Because, monthly mean DO
concentrations are not lower than the threshold value and monthly improvement are greater
than other options.
5.7: Limitation in quality modelling
A relatively limited number of chemical analysis data makes it difficult to simulate and to make more
detailed water quality modelling and analysis. The source data quality is not always clear due to
miscommunication and weak cooperation between water issues implementing organizations.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 65
Chapter 6: Conclusions and recommendations
6.1: Overall conclusions
A brief sequence of entire study as implemented is described as below:
− Proposal writing
− Organization and execution of field campaign
− Surface water quality assessment
− River flow model building and verification
− Water quality model building and analysis
− Conclusion and recommendations
A thesis proposal was written, based on real-time problems in Mongolia and mainly focussing on
water quality assessment, pollution mapping and modelling.
A field campaign was executed in the Ulaanbaatar city, Mongolia. Most vital information, data and
water samples were collected and analysed during the field work.
A method using a surface water quality index was utilized in assessment part. Six parameters i.e.
dissolved oxygen, ammonium, nitrate, nitrite, biochemical oxygen demand and chemical oxygen
demand are involved in index calculation. ILWIS 3.4 was used to visualise time series of thematic
water quality maps.
Before the surface water quality model development, a river flow model was developed using DMS
model code and using field observation. Totally, 24 chemical state variables and parameters took part
in quality model for dissolved oxygen.
More detailed specific conclusions are drawn at the end of chapters 3, 4 and 5. Overall conclusions,
recommendations, future research, output significance of study and limitations, difficulties, which
were encountered during study, are detailed in this chapter.
From this study, the following conclusions can be drawn:
− The general trend of water quality of the Tuul River showed a significant decrease throughout
the 11-year analysis period (1996-2006). Especially the downstream section of the Tuul
River, downstream of the Ulaanbaatar city, showed a decreasing trend in water quality. In the
year 1999, the water quality was comparatively good, with as main reason, new equipments
installation in the CWTP using JICA financial support. In the 2005, the Tuul River got
strongly polluted caused by poor operation of the CWTP.
− The water pollution of the Tuul River is gradually reducing along the downstream reach, but
not complete self-purification is reached at the Tuul-Altanbulag sampling point (35 km
down).
− Cold period negatively affects to the self-purification capacity and processes of the river.
Another reason is that discharges from pollutant sources are in this period greater than river
discharges.
− The CWTP remains the largest, strongest point pollution source in this branch of the Tuul
River. Political, scientific, engineering groups and water issues implementing organizations
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 66
have to pay more attention to the functioning of this treatment plant. Recent evidence
suggests that a main problem of this CWTP is reliable electricity supply.
− DMS software and existing quality model syntax can be used in this selected section of the
Tuul River.
− If real water level value or riverbed values are greater than one thousand, then DMS shown an
error message. Maybe, this is a bug of this software. This can be compensated by
substraction.
− Very limited number of chemical observations (once per month) can lead to significant
observation and model uncertainties. For example, sampled and analysed day and time is not
clear, only one analysis cannot always represent an entire month.
− When comparing the two models, a quality model (oxygen balance) is more sensitive than a
water flow and quantity model. Specially, the parameters β, Kd, Knit are very sensitive. Beta,
Kd, TKd, TKnit, KNDO have direct and Knit, SOD, fd have inverse relation to dissolved
oxygen concentration, respectively.
− Based on three developed scenarios, best one for SWQ improvement is first option. Because
of, monthly mean DO concentrations are not lower than the threshold value and monthly
improvement are greater than other options.
6.2: Recommendations
The pollution in the Tuul River is a serious problem, which needs to be addressed, urgently. A
different approach is suggested and needs to be taken in order to solve the complex issues involved.
� In Mongolia, many organizations are involved in water quantity and quality issues. Result of
that is miscommunication and data quality is unfavourable. In 2003, the government decided
to create a new organization, the Water Authority, which has the responsibility for water
resources issues. However, this organization still could not be involved in all urgent water
issues, such as the Tuul River water quality. Without any delay, the Water Authority should
take responsible for all urgent water issues in one centralized body and reorganize all other
small organizations under law.
� Another important issue is that a river basin management system needs to be developed for
the Tuul River basin. The NGIC project, which is funded by Dutch government, is trying to
develop such a management system. This project started in 2006 and will continue until 2009.
� The CWTP has to regularly renew its equipments (e.g. pumps, filtration, etc.), urgently and improve efficiency of the operation system, keep “good” controls in the operational system, constantly. In that case, maybe the CWTP needs to build its own independent electric power station.
� To implement a project for quality modelling, organizations in water issues should be involved. In the model, we should use datasets, which covers longer and real-time, sampled at higher time frequency and maybe also more sampling points. The wastewater carrying capacity of the Tuul River could then be estimated based on the model.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 67
6.3: Future research
The result of this study is not the end of this research. In future, we need to update some of the data
and improve the obtained results. For example:
− To change type of cross-section from trapezoid to asymmetric, and to improve the detailed
physical schematizations of the river network and pollution outfalls;
− To extend the entire study period in time;
− To collect and use more detailed data, that represents shorter time intervals and periods. This
means, chemical datasets, used in quality model, should be monitored (e.g. automatically) on
a daily basis rather than at a monthly scale. For more accurate result, we need to use shorter
time interval data;
− The spatial network could also be refined and more detailed, and the distances between nodes
could be decreased.
− Due to the elevation of the Ulaanbaatar city and the Tuul River, we should include air
pressure effects and water salinity data in oxygen saturation calculation.
− Although the Tuul River flows main origins are rainfall based and snowmelt streamflow, the
DMS model for both quantity and surface water quality could be set to interact with the
shallow groundwater using Moduflow option in DMS. Also could the DMS model be
extended with a rainfall-runoff of RAM component, to represent the whole water balance of
the river basin. This will require a future research and application project.
6.4: Output significance
Until now, this research has the following importance.
− The maps and assessment of river water contamination and a water quality model are useful
for decision makers and daily human life, well-being.
− In addition, outputs of research can be informative for urban planning.
− This research can be the basis of future research of river water contamination using different
spatial and temporal datasets.
6.5: Limitations of the study
In this study, the following limitations were faced:
− Some hydraulic and chemical datasets were very limited, generalized and not accessible by
internet. For example, cross-section data were not available. According to some information,
real cross-sectional data are not measured at Tuul-Altanbulag station.
− Most of data are costly and some of data still belonged to the national secret.
− The time to produce this MSc research was short and limited (only 6 months)
− Very limited number of chemical analysis data made it became difficult to simulate and to
make more detailed quality analysis. Data quality was not always clear due to
miscommunication and general weak cooperation between water issues implementing
organizations.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 68
− Correlation between water quality and quantity could not be studied in detail. Reason of that
is the Tuul River hydraulic parameters measurement and quality analysis are implemented by
two different organizations. Therefore, measurement dates and locations of both organizations
are not coinciding, which makes combined use very difficult. For example, daily discharge
measures in three different stations, that are belong to WS, NAMHEM, are situated in study
area. Ones in each month, the water quality analysis is implemented by CLEM in different
locations from the hydraulic measurements. Furthermore, the water quality analysis day of
sampling or observation is not written into the chemical dataset provided by CLEM.
− In the flow model simulation, use of actual water level data in both stations, especially at
Tuul-Altanbulag station, was sometimes confusing due to the implementation of water level
measurements using artificial reference surface level, instead of a permanent ground
reference.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 69
References
Adiyabadam, G. (1996), Chemical content of river, contamination rank and its model, National University of Mongolia, Ulaanbaatar.
Altansukh, O. (2000), Surface water pollution of Ulaanbaatar city, 43 pp, National University of Mongolia, Ulaanbaatar.
ASHRAE (2001), International Weather for Energy Calculations (IWEC weather files), edited, ASHRAE.
Baasandorj, D., and G. Davaa (2006), Tuul river basin of Mongolia - Integrated water resource management, First edition ed., Interpress, Co.,LTD, Ulaanbaatar.
Batima, P. (1998), Chemical content of Mongolian rivers and its quality assessment, National University of Mongolia, Ulaanbaatar.
Chapra, S. C. (1997), Surface water quality modelling, 843 pp., McGraw-Hill, Boston etc.
CSM (1998), Standard for surface water quality, edited, p. 30, Centre of Standardization and Measurement.
CWTP (2006), Annual report of CWTP, presentation, Ulaanbaatar.
Dashdeleg, N., and B. Bat (1971), Mongolian rivers, Ulaanbaatar.
Davaa, G., et al. (2006), Surface water of Mongolia, Environmental book of Mongolia, 55-82.
Dingman, S. L. (2002), Physical hydrology, Second edition ed., 646 pp., Prentice Hall, Upper Saddle River.
DMS (2004a), Duflow manual, edited, Duflow modelling studio.
DMS (2004b), User's guide, edited, Duflow Modelling System.
DMS (2007), Duflow Modelling System, edited, p. The official website of DMS, Delft, The Netherlands.
Dost, R. J. J. (2006), Water quality sampling and testing field guide, 1 ed., 136 pp., ITC, Enschede.
Edwards, K. (1998), Manning's Equation Calculator / Software, edited, LMNO Engineering, Research, and Software, Ltd.
Furniss, M., et al. (2006), Froude number and flow states, edited, USA.
Helsel, D. R., and R. M. Hirsch (1992), Statistical methods in water resources, 522 pp., Elsevier, Amsterdam etc.
Javzan, C., et al. (2004), Study of Tuul river contamination, Geo-ecology in Mongolia, 04, 213-219.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 70
Krenkel, P. A., and V. Novothy (1980), Water quality management, 671 pp., Academic Press, New York.
Lijklema, L., et al. (1996), Model for the DO in water system, in DMS, edited, Department of Water Quality Management and Aquatic Ecology, University of Wageningen, Wageningen.
Makkinga, A. A., et al. (1998), Modelling dissolved oxygen in the river Regge system using Duflow, European Water Management, 1(1998)1, pp. 37-42, 1, 6.
Mannaerts, C. (2007), The oxygen balance in rivers, in Water quality modelling, edited, Enschede, The Netherlands.
Merwade, V. (2005), Distributed flow routing, edited.
MNE (1997a), Causes of decreasing water resource of Tuul River and ways of its protection, Ministry of Nature and Environment, Ulaanbaatar.
MNE (1997b), Definition of quality grades used to surface water in Mongolia, edited, p. 14, Ministry of Nature and Environment.
MNE (2006), Report on State of Environment for 2004-2005, edited, p. 27, Ministry of Nature and Environment, Ulaanbaatar.
Munguntsetseg, A. (1987), Chemical content and contamination of Tuul river, National University of Mongolia, Ulaanbaatar.
Munguntsetseg, A., et al. (1982), Chemical component and pollution of Tuul River, National University of Mongolia, Ulaanbaatar.
NAMHEM (1999), Surface water in Mongolia, First edition ed., Interpress Co.,LTD, Ulaanbaatar.
Nash, J. E., and J. V. Sutcliffe (1970), River flow forecasting through conceptual models part I -- A discussion of principles, Journal of Hydrology, 10, 282-290.
Orchlon (1995), Pollution mitigating measures for the Tuul river - Three assessments, in Daily news, edited, Ulaanbaatar.
Orgil, B. (2007), Conceptual approach for water sub component of NGIC project, edited by M. o. N. a. Environment, p. 15, NGIC for NRM project.
Oyunchuluun, Y. (1999), Assessment of water quality using river bed insects, Pedagogical University of Mongolia, Ulaanbaatar.
Patankar, S. V. (1980), Numerical heat transfer and fluid flow, Taylor & Francis, [S.l.].
Rientjes, D. T. H. M. (2006), Modelling in hydrology, 250 pp., ITC, Department of Water Resources, Enschede, The Netherlands.
Rossiter, D. G. (2007), Using the R Environment for Statistical Computing, in Applied geo-statistics, edited, p. 44, International Institute for Geo-information Science & Earth Observation (ITC), Enschede.
Roza-Butler, N. (2004), An overview of the current condition of the Tuul river, Geo-ecology in Mongolia, 04, 220-226.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 71
Soninhishig, N. (1998), Algae in Tuul river, National University of Mongolia, Ulaanbaatar.
SRTM (2005), Shuttle Radar Topography Mission, edited, USA.
Thomann, R. V., and J. A. Mueller (1987), Principles of surface water quality modelling and control, Harper Collins, New York.
Tuvaanjav, G. (1972), Water regime of Tuul river, Journal for Mongolian geography, 11.
Tuvaanjav, G. (1978), Tuul River, hydrochemistry of its tributaries, Journal for Mongolian geography, 18.
Tuvaanjav, G. (1983), Changes of natural water quality and component near by cities, Journal for Mongolian geography, 23.
Uranbileg, L. (2000), River pollution in surrounding area of Ulaanbaatar city, in Conference for ecological problems in Mongolia, edited, Ulaanbaatar, National University of Mongolia.
White, F. M. (1998), Fluid Mechanics 4th Bk&Dk edition ed., 294 pp., Mcgraw-Hill College.
WHO (1996), Water quality assessments: A guide to the use of biota, sediments and water in environmental monitoring, Second edition ed., 626 pp., Spon, London etc.
Zagas, B. (1998), Assessment of water quality using water insects, National University of Mongolia, Ulaanbaatar.
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 73
Appendices
Appendix 1: Photos in the field
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 74
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 75
Appendix 2: Note for water field survey
A Information of sampling location
1 Name of state, province and municipality Mongolia, Ulaanbaatar,
2 Name of water body Tuul
3 Name of basin and sub-basin Tuul river basin and ……………………......... sub-basin
4 Sampled medium River
5 Sampling location
6 Geographical coordinate N(y)
E(x)
7 Elevation of sampling location ……………………………. meter a.s.l
8 Sampling site description
9 Water turbulence grade weak moderate high
10 Land use around sampling location
11 Possible contamination source
B Sample information 12 Name of collector Ochir Altansukh
13 Collected date and time 2007 ………………………………
14 Sample number or name
15 Sample container ……………… ml ……………..
16 Sampling method
standard method MNS ISO 5667-6 : 2001 “Water quality.
Sampling, 6th part. Guideline for sampling from river and
stream”
17 Depth of sampling middle of water depth
18 Water temperature …………….. 0C
19 Preservation method, if any
20 Identification of project MSc research project
21 Purpose of sampling water quality assessment
22 Weather condition during sampling
23 Any changes from normal weather
condition
24 Any specific observation in-situ
C Measurement result 25 Variable measured
26 Actual results
27 Measured unit
28 Analytical method
29 Measured place in-situ field laboratory standard laboratory
30 Name of instrument and laboratory
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 76
Appendix 3: Field measurements
NOTE OF FIELD SURVEY Measured by Mr. Ochir Altansukh (2007.09.17-22)
Geographical coordinate Measured variable
N River name Ex Ny
Elevation Water
turbulence 0C pH DO in mg l-1 EC in µµµµS cm-1
HACH file name
GPS point name
Photo file
1 Terelj - Terelj 47059/30.0// 107027/36.1// 1522 weak 7.9 7.59 8.73 49.6 Terelj 001 - 002 Ter - 01 01 - Terelj 01
2 Terelj - UB-2 tourist camp 47059/16.8// 107027/53.8// 1521 weak 9.3 7.47 7.78 52.4 Terelj 003 - 004 Ter - 02 02 - Terelj 02
3 Terelj - 03 47058/21.5// 107028/43.1// 1507 weak 10.1 7.37 7.71 51.5 Terelj 005 - 006 Ter - 03 none
4 Tuul - Uubulan 47048/26.4// 107022/50.3// 1383 weak 13.4 7.96 8.02 56.9 Terelj 007 - Tuul 001 T - 001 03 - Tuul 01
5 Tuul - Terelj bridge 47049/20.6// 107020/02.3// 1375 moderate 12.8 8.03 8.10 58.8 Tuul 002 - 003 T - 002 04 - Tuul 02
6 Tuul - 03 47049/43.5// 107019/25.0// 1372 moderate 12.7 8.04 8.07 59.3 Tuul 004 - 006 T - 003 05 - Tuul 03
7 Tuul - Nalaih 47049/14.0// 107015/56.4// 1364 weak 11.6 8.23 8.81 64.7 Tuul 007 - 008 T - 004 06 - Tuul 04
8 Nalaih - 01 47048/36.7// 107016/43.9// 1382 weak 16.1 7.36 6.15 768.0 Tuul 009 - 010 N - 001 07 - Nalaih 01
9 Nalaih - 02 47047/52.0// 107016/26.9// 1397 weak 18.9 8.57 7.43 778.0 Tuul 011 - 012 N - 002 08 - Nalaih 02
10 Nalaih - WWTS 47046/46.5// 107015/50.2// 1426 sewage 15.2 7.69 6.68 523.0 Tuul 013 - 014 N - 003 09 - WWTS 01
11 Nalaih - Spring 47046/46.5// 107015/50.2// 1426 spring water 14.3 8.51 8.85 1206.0 Tuul 017 - 018 N - 003 10 - Nalaih 03
12 Confluence of Nalaih spring and sewage 47046/50.6// 107015/51.8// 1422 weak 16.3 7.95 5.27 580.0 Tuul 019 - 020 N - 004 none
13 Tuul -05 47053/06.7// 107014/50.5// 1350 moderate 13.2 8.54 8.45 280.0 Tuul 021 - 022 T - 005 11 - Tuul 05
14 Tuul -06 47053/45.2// 107012/50.3// 1352 moderate 12.9 8.21 8.31 398.0 Tuul 023 - 024 T - 006 12 - Tuul 06
15 Tuul - Gachuurt 47054/28.1// 107010/40.9// 1340 weak 12.0 8.15 8.24 132.0 Tuul 025 - 026 T - 007 13 - Tuul 07
16 Tuul - Khar usan tohoi 47054/39.9// 107007/44.0// 1334 moderate 9.3 7.95 8.88 244.0 Tuul 027 - 028 T - 008 14 - Tuul 08
17 Tuul - Bayanzurh bridge (upstream) 47053/55.1// 107004/12.8// 1313 moderate 9.4 8.06 8.81 165.1 Tuul 029 - 030 T - 009 15 - Tuul 09
18 Tuul - Bayanzurh bridge (downstream) 47053/28.1// 107003/04.7// 1309 weak 11.5 8.78 9.21 147.8 Tuul 031 - 032 T - 010 16 - Tuul 10
19 Tuul - Railway bridge (downstream) 47053/36.7// 107001/55.0// 1305 weak 14.0 7.98 7.91 335.0 Tuul 033 - 035 T - 011 17 - Tuul 11
20 Tuul - 12 47053/30.5// 106059/45.2// 1297 weak 15.8 8.31 8.57 223.0 Tuul 036 - 037 T - 012 18 - Tuul 12
21 Tuul - Zaisan 47053/19.4// 106055/05.7// 1306 weak 15.0 8.57 8.23 115.8 Tuul 038 - 039 T - 013 19 - Tuul 13
22 Tuul - 14 47053/26.6// 106052/48.6// 1276 moderate 15.4 8.48 8.40 134.2 Tuul 040 - 041 T - 014 20 - Tuul 14
23 Tuul - Niseh bridge (downstream) 47053/17.4// 106050/47.3// 1276 weak 14.9 8.37 8.18 231.0 Tuul 042 - 043 T - 015 21 - Tuul 15
24 Selbe - Khandgait 48007/03.7// 106053/27.7// 1497 weak 9.5 8.09 8.70 224.0 Tuul 044 - 045 S - 001 22 - Selbe 01
25 Khandgait - 01 48006/44.7// 106054/07.4// 1505 weak 12.1 8.10 8.74 309.0 Selbe 001 - 002 S - 002 23 - Khandgait 01
26 Selbe - Belh bridge 48004/40.3// 106053/49.1// 1443 weak 13.2 8.37 8.25 267.0 Selbe 003 - 004 S - 003 24 - Selbe 02
27 Selbe - Sharga Morit 48002/58.7// 106054/10.7// 1411 moderate 14.4 8.67 8.40 249.0 Selbe 005 - 006 S - 004 25 - Selbe 03
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 77
28 Selbe - Shadivlan 48001/00.2// 106055/16.9// 1381 weak 15.9 8.61 8.01 310.0 Selbe 007 - 008 S - 005 26 - Selbe 04
29 Selbe - Dambadarjaa bridge 47058/36.8// 106055/32.9// 1346 weak 17.9 9.50 8.45 282.0 Selbe 009 - 010 S - 006 27 - Selbe 05
30 Selbe - Selbe bridge 47055/44.3// 106055/53.3// 1303 weak 18.1 9.06 9.98 424.0 Selbe 011 - 012 S - 007 28 - Selbe 06
31 Selbe - Lion bridge 47055/05.0// 106055/45.9// 1304 weak 18.5 9.13 9.41 508.0 Selbe 013 - 014 S - 008 29 - Selbe 07
32 Mouth of Tolgoit stream 47053/51.6// 106049/46.8// 1280 weak 17.9 8.24 10.06 678.0 Selbe 015 - 016 To - 001 30 - Tolgoit 01
33 Tuul - Sonsgolon 47052/44.6// 106048/01.8// 1256 moderate 16.3 8.80 7.99 110.6 Tuul 046 - 047 T - 016 31 - Tuul 16
34 Confluence of Tuul and Selbe 47053/01.7// 106048/47.5// 1266 weak 16.3 8.77 7.76 320.0 Tuul 048 - 049 T - 017 32 - Tuul 17
35 Tuul - 18 47053/01.7// 106048/47.5// 1266 weak 16.2 8.76 7.80 75.6 Tuul 050 - 051 T - 017 33 - Tuul 18
36 Mouth of Selbe stream 47053/01.7// 106048/47.5// 1266 weak 16.0 8.76 8.27 527.0 Selbe 017 - 018 T - 017 34 - Selbe 08
37 Selbe - 09 47053/27.9// 106049/19.5// 1261 weak 15.9 8.30 3.54 540.0 Selbe 019 - 020 S - 009 35 - Selbe 09
38 Confluence of Selbe and Tolgoit 47053/47.1// 106049/42.7// 1270 weak 15.9 8.32 9.69 690.0 Selbe 021 - 022 S - 010 36 - Selbe 10
39 Selbe - 11 47053/47.1// 106049/42.7// 1270 weak 13.6 7.38 2.82 418.0 Selbe 023 - 024 S - 010 37 - Selbe 11
40 Tuul - Sonsgolon bridge (downstream) 47052/28.7// 106046/50.1// 1262 moderate 12.0 8.26 8.70 77.6 Tuul 052 - 053 T - 018 38 - Tuul 19
41 Tuul - Niseh 47052/01.6// 106043/05.9// 1246 moderate 13.1 8.41 8.74 468.0 Tuul 054 - 055 T - 019 39 - Tuul 20
42 WWTP - Railway bridge 47052/13.5// 106041/35.6// 1239 weak 17.8 7.73 0.62 1134.0 Tuul 056 - 057 T - 020 40 - WWTS 02
43 Tuul - Songino resort (upper) 47052/01.8// 106041/42.8// 1238 moderate 13.6 8.39 8.64 81.0 Tuul 058 - 059 T - 021 41 - Tuul 21
44 Confluence of Tuul and WWTP 47052/01.8// 106041/42.8// 1238 weak 14.5 7.96 7.21 213.0 Tuul 060 - 061 T - 021 none
45 Confluence of Tuul and WWTP (30 m in downstream) 47051/52.0// 106041/32.2// 1226 weak 15.6 7.82 6.42 281.0 Tuul 062 - 063 T - 022 42 - Tuul 22
46 Tuul - Bio bridge 47050/51.7// 106040/29.7// 1235 moderate 13.9 7.95 7.19 242.0 Tuul 064 - 065 T - 023 43 - Tuul 23
47 Tuul - Songino (upper) 47051/17.8// 106041/23.2// 1237 weak 14.3 8.50 9.05 135.2 Tuul 066 - 067 T - 024 44 - Tuul 24
48 Discharge from WWTS of Bio 47050/39.0// 106040/12.7// 1248 discharge 16.1 7.48 6.14 177.2 Tuul 068 - 069 T - 025 45 - WWTS 03
49 Tuul - Chicken farm (upstream) 47046/21.0// 106035/59.2// 1211 weak 15.7 8.41 8.13 238.0 Tuul 070 - 071 T - 026 46 - Tuul 25
50 Tuul - Khadanhyasaa (Salhitiin Gatsaa) 47045/08.9// 106030/02.6// 1205 weak 9.4 7.72 6.75 422.0 Tuul 072 - 073 T - 027 47 - Tuul 26
51 Tuul - Altanbulag 47042/59.8// 106024/29.5// 1180 weak 10.6 7.58 7.41 305.0 Tuul 074 - 075 T - 028 48 - Tuul 27
52 Tuul - Altanbulag (Guurnii Gatsaa) 47041/53.4// 106017/40.6// 1185 weak 16.0 8.76 10.05 274.0 Tuul 076 - 077 T - 029 49 - Tuul 28
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 78
Appendix 4: Hydro-chemical dataset
Chemical monthly data of the Tuul river and its tributaries
EC DO NH4+ NO2
- NO3-
Total mineral N
COD-Mn BBOODD Date pH
µS cm -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1
Tuul - Songino (down) January-96 7.5 571.0 7.37 15.30 0.066 1.42 16.786 6.5 14.0
February-96 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-96 8.6 389.0 5.93 15.80 0.078 1.77 17.648 11.1 14.3
April-96 7.8 590.0 3.95 1.04 0.149 0.88 2.069 3.6 May-96 7.8 174.4 7.90 1.18 0.042 0.69 1.912 3.7 4.0
June-96 8.0 268.0 9.42 2.12 0.063 1.28 3.463 3.5 0.2 July-96 7.7 135.3 8.78 0.29 0.022 0.50 0.812 4.0 6.8
August-96 7.4 135.9 9.24 0.51 0.034 0.32 0.864 4.1 2.9 September-96 7.6 164.7 8.26 1.54 0.040 0.47 2.050 2.4 1.6
October-96 7.6 237.0 8.06 0.83 0.037 0.66 1.527 2.9 4.3 November-96 7.6 504.0 3.12 11.40 0.151 2.13 13.681 4.4 10.7 December-96 n.a n.a n.a n.a n.a n.a n.a n.a n.a
Annual average -96 7.8 316.9 7.20 5.00 0.068 1.01 6.081 4.6 6.5 January-97 7.9 565.0 2.61 n.a n.a n.a n.a 3.7 12.46<
February-97 8.0 545.0 2.13 18.90 0.069 0.96 19.929 5.3 37.7 March-97 7.9 437.0 4.10 0.925 0.925 n.a 3.5
April-97 7.7 207.0 8.16 1.60 0.047 0.69 2.337 4.1 7.3 May-97 7.7 270.0 8.16 3.94 0.108 0.81 4.858 2.9 8.16<
June-97 n.a n.a n.a n.a n.a n.a n.a n.a n.a July-97 7.9 96.6 10.75 0.24 0.032 0.41 0.682 4.8 3.1
August-97 7.7 81.9 8.06 0.50 0.014 0.514 4.0 1.7 September-97 7.9 144.7 9.36 1.11 0.024 0.29 1.424 2.6 2.9
October-97 8.0 227.0 9.07 2.39 0.079 0.92 3.389 3.5 6.9 November-97 8.3 399.0 4.68 8.92 0.187 1.02 10.127 4.1 15.56< December-97 8.5 456.0 4.97 12.70 0.135 1.71 14.545 5.3 18.7
Annual average -97 7.9 311.7 6.55 5.59 0.162 0.85 5.873 4.0 10.2 January-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a
February-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-98 8.7 502.0 n.a 9.59 0.313 1.43 11.333 3.6 20.8
April-98 7.2 n.a 9.09 2.24 0.832 2.28 5.352 5.4 13.4 May-98 7.0 217.0 9.16 2.28 0.250 0.63 3.160 3.2 4.8
June-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a July-98 7.7 109.5 7.37 1.05 0.035 0.23 1.315 3.0 2.7
August-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a September-98 7.3 85.2 8.74 0.28 0.007 0.57 0.857 5.0 1.9
October-98 7.3 168.9 11.32 0.67 0.175 0.56 1.405 3.1 6.6 November-98 6.7 330.0 9.42 3.26 0.003 1.09 4.353 4.9 7.0 December-98 7.4 360.0 6.69 10.20 0.124 1.92 12.244 5.9 8.06<
Annual average -98 7.4 253.2 8.83 3.70 0.217 1.09 5.002 4.3 8.2 January-99 n.a n.a n.a n.a n.a n.a n.a n.a n.a
February-99 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-99 7.8 470.0 9.27 11.74 0.028 0.89 12.658 5.5 5.4
April-99 7.3 63.8 10.84 0.46 0.007 0.20 0.667 6.6 2.0 May-99 7.9 110.4 11.08 0.81 0.020 0.31 1.140 3.9 4.8
June-99 6.9 63.4 8.27 0.32 0.010 0.30 0.630 7.3 3.0 July-99 8.1 80.7 7.80 0.30 0.110 0.18 0.590 4.6 1.4
August-99 8.1 60.8 8.59 0.11 0.021 0.09 0.221 3.0 2.0 September-99 8.0 115.0 11.05 0.69 0.014 0.30 1.004 3.2 1.4
October-99 7.9 212.0 10.41 2.36 0.047 0.72 3.127 4.9 4.8 November-99 8.0 190.8 9.22 2.80 0.032 0.74 3.572 1.8 3.0 December-99 n.a n.a n.a n.a n.a n.a n.a n.a n.a
Annual average -99 7.8 151.9 9.61 2.18 0.032 0.41 2.623 4.5 3.1 January-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a
February-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a
April-00 8.0 250.0 8.80 4.25 0.169 1.41 5.829 4.5 7.9 May-00 7.4 244.0 6.60 3.67 0.070 0.41 4.150 13.0 6.6
June-00 8.2 195.4 8.70 2.52 0.073 0.80 3.393 3.2 5.1 July-00 8.3 141.0 7.90 1.05 0.034 0.35 1.434 2.5 2.3
August-00 8.1 132.0 8.10 1.10 0.042 0.48 1.622 2.0 2.9 September-00 7.9 107.0 10.50 0.74 0.013 0.23 0.983 1.9 2.8
October-00 7.6 349.0 9.50 6.26 0.216 0.72 7.196 6.6 8.3 November-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a December-00 7.6 617.0 6.98 11.82 0.219 1.01 13.049 5.5 12.7
Annual average -00 7.9 254.4 8.39 3.93 0.105 0.68 4.707 4.9 6.1 January-01 8.0 563.0 4.47 13.06 0.282 1.06 14.402 8.3 10.8
February-01 7.6 543.0 8.62 14.00 0.322 3.65 17.972 5.7 13.6 March-01 n.a 685.0 7.01 18.52 0.167 1.52 20.207 6.6 6.1
April-01 7.9 195.0 9.23 1.66 0.011 0.47 2.141 12.3 6.0 May-01 6.9 122.0 8.18 1.42 0.008 0.33 1.758 8.2 3.3
June-01 n.a 199.0 5.68 0.09 0.003 3.00 3.093 4.6 5.7
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 79
July-01 7.2 207.0 7.18 3.13 0.679 0.36 4.169 4.9 6.3 August-01 7.3 115.0 7.85 1.12 0.048 0.19 1.358 5.1 4.6
September-01 7.6 173.0 7.90 1.78 0.025 1.30 3.105 4.0 4.3 October-01 8.0 300.0 7.30 4.53 0.049 0.57 5.149 6.1 7.3
November-01 8.0 681.0 2.74 2.55 0.010 0.11 2.670 18.5 29.6 December-01 n.a n.a n.a n.a n.a n.a n.a n.a n.a
Annual average -01 7.6 343.9 6.92 5.62 0.146 1.14 6.911 7.7 8.9 January-02 7.9 716.0 3.34 17.42 0.015 0.03 17.465 5.8 33.5
February-02 7.9 713.0 2.58 17.29 0.138 0.63 18.058 13.7 41.0 March-02 7.8 n.a 1.79 26.78 0.067 0.02 26.867 5.7 48.6
April-02 7.5 723.0 3.33 21.44 0.005 0.07 21.515 5.8 17.3 May-02 7.9 86.0 8.06 22.95 0.013 0.35 23.313 4.9 3.5
June-02 8.4 168.7 9.59 1.74 n.a 0.39 n.a 4.8 9.1 July-02 7.8 109.8 8.43 0.67 0.156 0.18 1.006 2.5 5.0
August-02 7.4 228.0 5.71 0.01 0.047 0.39 0.447 3.3 4.8 September-02 7.8 271.0 10.06 3.18 0.075 0.31 3.565 3.1 2.9
October-02 8.1 381.0 8.36 5.57 0.025 0.80 6.395 4.9 6.0 November-02 8.4 770.0 3.54 16.17 0.120 0.53 16.820 11.0 9.5 December-02 n.a n.a n.a n.a n.a n.a n.a n.a n.a
Annual average -02 7.9 416.7 5.89 12.11 0.066 0.34 13.545 5.9 16.4 January-03 7.7 605.0 4.76 14.85 0.040 0.44 15.330 20.1 29.3
February-03 8.4 862.0 3.43 19.73 0.020 0.07 19.820 21.1 38.8 March-03 8.1 789.0 4.17 17.18 0.088 0.17 17.438 13.0 15.7
April-03 6.8 164.4 7.75 0.62 2.579 0.14 3.339 16.1 15.7 May-03 6.9 103.5 8.57 0.78 0.026 0.28 1.086 5.6 2.7
June-03 7.3 196.5 6.85 2.42 0.027 0.34 2.787 2.8 6.1 July-03 6.8 94.3 5.90 0.49 0.001 0.23 0.721 4.1 2.0
August-03 6.8 90.0 7.68 0.46 0.008 0.60 1.068 4.1 1.7 September-03 7.2 0.2 11.84 1.57 0.007 0.40 1.977 3.4 3.4
October-03 7.4 0.3 10.32 3.28 0.020 0.05 3.350 7.0 10.0 November-03 n.a n.a n.a n.a n.a n.a n.a n.a n.a December-03 8.0 0.98 18.16 0.017 0.08 18.257 n.a n.a
Annual average -03 7.4 290.5 6.57 7.23 0.258 0.25 7.743 9.7 12.5 January-04 7.2 540.0 4.97 16.53 0.038 0.79 17.358 13.0 29.3
February-04 7.3 635.0 4.89 13.04 0.070 1.13 14.240 14.0 31.1 March-04 8.1 735.0 9.32 16.14 0.032 1.04 17.212 10.0 26.5
April-04 6.5 168.8 9.80 0.63 0.002 0.55 1.182 14.0 5.6 May-04 7.4 118.4 9.74 1.29 0.016 0.27 1.576 7.4 4.0
June-04 7.2 150.7 5.71 1.38 0.042 0.28 1.702 3.3 3.5 July-04 6.9 98.2 9.07 0.05 0.023 0.02 0.093 3.4 4.0
August-04 7.2 149.2 8.90 1.39 0.037 0.04 1.464 4.5 8.6 September-04 7.8 113.2 9.49 0.75 0.021 0.02 0.792 4.0 5.4
October-04 7.9 175.4 11.64 1.98 0.024 0.03 2.034 4.1 8.9 November-04 7.5 599.0 1.15 10.50 0.002 0.09 10.587 15.0 64< December-04 7.4 493.0 0.82 14.41 0.014 0.31 14.734 39.5 98.5
Annual average -04 7.4 331.3 7.13 6.51 0.027 0.38 6.915 11.0 20.5 January-05 8.2 873.0 1.25 12.00 0.007 0.09 12.097 26.7 55.3
February-05 7.7 613.0 8.28 13.43 0.045 1.54 15.015 13.2 41.0 March-05 7.9 753.0 0.90 1.40 0.005 n.a n.a 37.0 62.8
April-05 7.2 712.0 2.61 12.62 0.325 0.24 13.185 39.0 37.4 May-05 7.9 141.9 8.39 1.64 0.074 0.32 2.034 5.0 2.4
June-05 7.2 134.1 9.28 1.04 0.050 0.15 1.240 6.6 6.6 July-05 7.2 161.7 8.48 0.94 0.045 1.11 2.095 4.4 6.6
August-05 8.3 141.1 9.76 1.21 0.031 1.49 2.731 4.2 5.1 September-05 8.2 116.1 10.68 0.99 0.009 0.08 1.079 3.4 2.8
October-05 7.9 236.0 10.16 5.94 0.048 0.35 6.338 5.2 5.1 November-05 7.9 645.0 4.00 14.32 0.048 0.54 14.908 29.6 31.6 December-05 7.7 649.0 2.40 13.33 0.094 0.56 13.984 20.4 51.6
Annual average -05 7.8 431.3 6.35 6.57 0.065 0.59 7.701 16.2 25.7 January-06 7.7 719.0 4.00 18.62 0.127 1.90 20.647 23.6 33.1
February-06 7.5 773.0 n.a 21.76 0.045 1.69 23.495 n.a 55.4 March-06 8.2 721.0 0.77 20.29 0.339 0.09 20.719 38.4 44.2
April-06 7.3 695.0 0.46 7.12 0.084 0.19 7.394 30.4 43.8 May-06 6.9 94.1 8.62 0.37 0.010 0.44 0.820 12.0 2.3
June-06 6.9 104.7 7.85 0.92 0.028 0.16 1.108 6.0 4.8 July-06 7.6 120.5 7.55 0.53 0.065 0.36 0.955 4.5 1.3
August-06 7.4 171.5 8.32 0.51 0.191 1.29 1.991 7.6 5.2 September-06 7.5 138.0 7.61 1.56 0.050 0.69 2.300 7.2 4.7
October-06 7.0 406.0 4.44 8.32 0.122 8.442 34.7 20.6 November-06 7.6 789.0 5.98 12.16 0.165 0.87 13.195 14.5 20.6 December-06 7.7 876.0 4.70 21.78 0.085 0.63 22.495 27.8 39.7
Annual average -06 7.4 467.3 5.48 9.50 0.109 0.76 10.297 18.8 23.0 September-07 8.0 242.0 7.19 6.41 0.112 0.67 7.204 8.7 13.7
Hint: in last row, September 2007, pH, EC, DO measured during MSc field survey and samples for NH4
+, NO2-, NO3
-, COD-Mn, BOD5 collected by myself, analyzed in CLEM.
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 80
Appendix 5: Hydraulic dataset
Hydraulic measurement of the Tuul River, Ulaanbaatar station in 2006 Source: NAMHEM, Water sector
Water velocity by m s-1 Water depth by m N Date Direction
Water quality in direction
Water level by m (H)
Water discharge by m3 s-1 (Q)
Water area by m2
(F) mean max
River width by m mean max
Water surface gradient by %
Measured method
Analyzed method
1 21-Apr-06 1 very polluted 75 2.1 6.16 0.34 0.6 48.6 0.13 0.38 гр-21 4/4 задлаг
2 4-May-06 1 clean 131 25 30.3 0.82 1.37 64 0.47 0.8 гр-21 6/11 задлаг
3 12-May-06 1 clean 94 6.69 14.1 0.47 0.94 51.9 0.27 0.49 гр-21 5/7 задлаг
4 18-May-06 1 clean 142 41.2 39 1.06 1.7 66.5 0.59 0.92 гр-21 6/11 задлаг
5 24-May-06 1 clean 137 32.8 35.4 0.93 1.56 65.6 0.54 0.86 гр-21 6/11 задлаг
6 28-May-06 1 clean 120 13.1 21.5 0.61 0.9 58.5 0.37 0.55 гр-21 4/9 задлаг
7 1-Jun-06 1 clean 130 29 33.2 0.87 1.54 64.2 0.52 0.82 3.9 гр-21 6/10 задлаг
8 4-Jun-06 1 clean 178 69.9 64.7 1.08 1.38 buoy задлаг
9 6-Jun-06 1 clean 212 186.4 86.1 2.36 3.63 buoy задлаг
10 7-Jun-06 1 clean 197 124 78.9 1.57 2.1 buoy задлаг
11 9-Jun-06 1 clean 158 42 51 0.82 1.38 buoy задлаг
12 21-Jun-06 1 clean 124 21.5 28.3 0.76 1.33 61.9 0.46 0.75 4.1 гр-21 6/11 задлаг
13 24-Jun-06 1 clean 118 15.4 24.8 0.62 1.13 59.4 0.42 0.69 4.1 гр-21 6/10 задлаг
14 30-Jun-06 1 clean 146 45.2 41.1 1.1 1.78 67.2 0.61 0.95 4 гр-21 6/11 задлаг
15 1-Jul-06 1 clean 149 49.4 43 1.15 1.85 67.7 0.64 0.98 3.1 гр-21 6/11 задлаг
16 9-Jul-06 1 clean 172 53.8 57.8 0.93 1.25 buoy задлаг
17 10-Jul-06 1 clean 200 136 78.9 1.72 2.22 buoy задлаг
18 12-Jul-06 1 clean 190 106 71.8 1.48 1.9 buoy задлаг
19 17-Jul-06 1 clean 138 38.2 37.7 1.01 1.67 66 0.57 0.9 3 гр-21 6/11 задлаг
20 25-Jul-06 1 clean 125 22.2 28.7 0.77 1.32 62 0.46 0.75 3 гр-21 6/10 задлаг
21 31-Jul-06 1 clean 132 32.6 34.7 0.94 1.61 65 0.53 0.85 3 гр-21 6/11 задлаг
22 1-Aug-06 1 clean 147 48.2 42.6 1.13 1.79 67.6 0.63 0.98 4 гр-21 6/11 задлаг
23 3-Aug-06 1 clean 134 34 35.4 0.96 1.64 65.2 0.54 0.86 3.82 гр-21 6/11 задлаг
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 81
24 12-Aug-06 1 clean 120 17.4 26 0.67 1.15 60.1 0.43 0.71 3.72 гр-21 6/10 задлаг
25 15-Aug-06 1 clean 112 12.7 22.5 0.56 1.04 58.3 0.39 0.65 3.72 гр-21 6/9 задлаг
26 31-Aug-06 1 clean 107 10.2 20.4 0.5 0.92 57.4 0.36 0.63 3.63 гр-21 6/10 задлаг
27 4-Sep-06 1 clean 105 9.26 19.4 0.48 0.91 57.2 0.34 0.59 3.53 гр-21 6/10 задлаг
28 10-Sep-06 1 clean 108 10.5 20.8 0.5 0.94 57.6 0.36 0.62 3.63 гр-21 6/9 задлаг
29 15-Sep-06 1 clean 103 7.06 18.2 0.39 0.85 56.7 0.32 0.57 3.53 гр-21 6/8 задлаг
30 26-Sep-06 1 clean 110 11.6 21.9 0.53 1 58 0.38 0.64 3.63 гр-21 6/9 задлаг
31 30-Sep-06 1 clean 113 13.4 23.5 0.57 1.06 58.6 0.4 0.67 3.63 гр-21 6/9 задлаг
32 2-Oct-06 1 clean 111 12.9 22.7 0.57 1.07 58.3 0.39 0.64 3.43 гр-21 6/9 задлаг
33 8-Oct-06 1 clean 106 9.63 19.8 0.49 0.91 57.3 0.35 0.61 3.43 гр-21 6/9 задлаг
34 14-Oct-06 1 clean 100 7.13 16.7 0.43 0.83 54.7 0.3 0.55 3.43 гр-21 6/8 задлаг
35 29-Oct-06 1 clean 92 4.18 12.4 0.34 0.73 49.7 0.25 0.48 3.43 гр-21 6/8 задлаг
36 2-Nov-06 1 very polluted 96 6.11 14.6 0.42 0.82 52.5 0.28 0.52 3.43 гр-21 6/8 задлаг
37 10-Nov-06 1 very polluted 106 2.46 12 0.2 0.37 54.8 0.31 0.56 1.96 гр-21 5/5 задлаг
38 19-Nov-06 1 very polluted 100 0.66 7.4 0.09 0.16 55.9 0.29 0.38 1.67 гр-21 4/4 задлаг
39 30-Nov-06 1 very polluted 103 0.13 3.4 0.04 0.06 56.1 0.21 0.56 1.76 гр-21 2/2 задлаг
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 82
Appendix 6: CWTP chemical dataset
Month
Sam
plin
g po
int
Vol
ume,
tho
usan
d m
3
Wat
er
tem
pera
ture
, 0 C
pH
Tur
bidi
ty, c
m
Rem
aini
ng
capa
city
, ml
Susp
ende
d so
lid,
mg
l-1
BO
D5,
mg
l-1
BO
D20
, mg
l-1
DO
, mg
l-1
CO
D, m
g l-1
Alk
alin
ity,
g-e
cv l-1
Chl
orid
e, m
g l-1
NH
4+ , m
g l-1
NO
3- , mg
l-1
NO
2- , mg
l-1
PO
4-2, m
g l-1
Tot
al P
, mg
l-1
SO4-2
, mg
l-1
Tot
al C
r, m
g l-1
Tot
al N
, mg
l-1
EC
Col
or
incoming 14.00 8.76 2.20 3.17 289.00 301.50 450.00 859.03 4.20 71.02 30.18 4.76 1.25 3.13 84.79 1.30 28.30 614.00 5.07 January
discharging 4580
11.70 7.30 4.50 0.23 47.66 53.95 65.30 2.42 135.54 5.37 71.44 26.54 5.41 0.08 2.01 44.46 0.22 30.56 664.00 5.03
incoming 11.50 8.67 2.25 2.75 261.00 271.94 323.40 505.44 3.65 47.86 29.00 5.53 0.47 1.48 33.65 0.12 56.04 419.50 5.20 February
discharging 4219
9.00 7.25 11.25 33.00 38.60 53.23 4.82 112.32 4.26 43.4 16.79 7.81 0.08 1.14 25.08 0.02 16.81 407.50 4.55
incoming 10.00 8.49 2.26 2.93 461.61 257.74 360.00 737.71 4.80 103.13 30.36 1.35 0.33 1.40 58.21 0.64 55.48 615.33 5.17 March
discharging 4702
8.67 7.01 10.40 35.33 30.60 42.52 4.41 139.10 4.07 60.52 29.33 3.21 0.01 0.81 35.43 0.16 30.26 502.33 4.63
incoming 14.00 7.45 2.43 2.33 377.33 243.21 329.00 603.72 3.89 79.84 33.59 1.86 0.15 2.16 30.95 0.54 23.82 608.00 5.73 April
discharging 4896
13.00 7.01 6.30 39.66 24.13 37.00 2.3 140.51 3.55 61.29 18.72 3.66 0.01 1.40 22.09 0.35 14.29 523.67 4.90
incoming 13.00 7.47 3.00 1.53 257.30 257.10 430.00 332.27 3.95 71.03 19.93 3.22 0.03 0.81 1.50 0.77 534.30 5.70 May
discharging 4532
12.00 7.13 15.90 18.00 17.89 41.40 3.54 83.03 3.53 57.18 16.50 4.34 0.01 0.58 1.33 0.06 486.70 4.50
incoming 17.00 6.84 2.55 2.25 247.50 300.50 375.00 582.88 3.55 62.88 23.82 3.68 0.00 0.30 38.80 0.31 547.00 4.60 June
discharging 4233
15.00 6.85 10.20 15.50 33.20 40.80 1.99 111.02 3.63 46.93 15.74 4.94 0.00 0.13 27.84 0.00 517.50 4.10
incoming 19.00 6.62 2.75 3.25 297.67 325.65 496.99 3.97 72.58 20.08 7.07 0.06 1.55 1.15 37.33 0.18 600.00 5.17 July
discharging 4330
17.00 6.71 17.25 38.00 30.99 1.87 85.70 3.39 53.56 21.50 8.00 0.00 1.52 0.63 24.80 0.16 558.00 4.67
incoming 20.00 6.92 2.00 3.96 405.00 284.22 530.14 4.41 96.08 29.69 4.49 0.36 2.51 38.03 0.14 687.00 5.30 August
discharging 4550
18.00 6.83 7.40 25.30 29.12 1.80 86.98 3.96 71.13 21.30 5.41 0.05 1.92 21.48 0.00 685.00 5.50
incoming 18.00 7.68 2.73 3.20 288.00 233.93 457.68 4.96 109.75 24.25 2.31 0.10 2.23 64.00 0.44 667.50 5.40 September
discharging 4790
16.00 6.98 7.90 63.60 33.80 1.39 95.55 4.24 92.04 21.01 3.63 0.09 1.31 50.73 0.17 703.70 7.43
incoming 18.00 7.95 3.07 3.03 529.33 264.77 552.96 5.01 101.68 27.08 2.68 1.09 2.86 68.28 0.21 645.00 5.73 October
discharging 5188
16.00 6.70 11.37 67.50 32.14 118.59 4.05 78.03 16.90 5.08 0.03 1.12 32.42 0.05 638.67 5.07
incoming 14.00 8.10 2.80 3.50 406.67 297.27 482.98 4.43 92.17 24.17 3.64 1.11 2.68 3.41 0.48 681.33 8.23 November
discharging 4849
13.00 6.92 11.00 45.33 36.03 6.93 78.98 3.94 73.53 19.37 4.54 0.05 1.00 2.27 0.12 612.33 5.45
incoming 14.00 8.67 2.60 3.33 468.00 345.35 267.36 5.39 83.84 23.17 2.83 1.57 3.13 3.00 0.93 595.33 6.20 December
discharging 5215
11.30 7.13 7.50 48.00 43.35 126.72 4.15 67.99 15.55 5.33 0.03 1.70 2.41 0.09 621.33 5.17
incoming 15.21 7.80 2.55 2.94 357.37 281.93 534.10 4.35 82.66 26.28 3.62 0.54 2.02 2.27 50.45 0.51 40.91 601.19 5.63 Mean
discharging 4674
13.39 6.99 10.08 0.02 39.74 33.65 3.15 109.50 4.01 64.75 19.94 5.11 0.04 1.22 1.66 31.59 0.12 22.98 576.73 5.08
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 83
Appendix 7: A flow model setup
Upper CS – Zaisan Trapezoid Floor width=48.6 Slope=5.71 Max depth=1.5 Floor level=285.25 Surface level=292.6 Sonsgolon CS Trapezoid Floor width=44.5 Slope=5.65 Max depth=1.5 Floor level=265.44 Surface level=272.12 Songino (upper) CS Trapezoid Floor width=41.3 Slope=5.61 Max depth=1.5 Floor level=249.6 Surface level=255.73 Songino (down) CS Trapezoid Floor width=41.0 Slope=5.6 Max depth=1.5 Floor level=247.62 Surface level=253.69 Chicken farm CS Trapezoid Floor width=36.8 Slope=5.54 Max depth=1.5 Floor level=227.81 Surface level=233.2 Khadanhyasaa CS Trapezoid Floor width=33.5 Slope=5.5 Max depth=1.5 Floor level=211.97 Surface level=216.82 Altanbulag CS Trapezoid Floor width=26.6 Slope=5.4 Max depth=1.5 Floor level=178.3 Surface level=182 Lower CS – Extension Trapezoid Floor width=10 Slope=20 Max depth=1.5 Floor level=173.3 Surface level=177
BC upstream – Zaisan Q Starting
05.3.1-06.12.1 Q=0.033 H = 286.44=285.25+1.19 Location: 643389 E, 5305700 N
BC midstream - none Calibration point using Q at Altanbulag ≈ 54 km far from upstream
BC downstream – Artificial Q Starting
05.3.1-06.12.1 Q=-1.38
10 days constant H = 174.8=173.3+1.5
Location: 587593 E, 5301654 N 20 km far from Altanbulag
Section - Tuul ≈ 54 km long CS interpolated Min distance 10 m and max 200
m n=0.06
DP – CWTP Q Starting
05.3.1-06.12.1 Q=1.814 Location: 17.5 km far from upstream
DP – Bio Q Starting
05.3.1-06.12.1 Q=0.0172 Location: 2.5 km far from CWTP
Initial condition Q 0.198 1.38 H 286.44 174.8
Calculation Starting
05.3.1-06.12.1 Time step=1” Theta=0.9 Distance=200 Total
Manning Summer: 4.1 – 12.1
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 84
Appendix 8: Manning value
Manning's n for Channels (Chow, 1959)
Type of Channel and Description Minimum Normal Maximum
Natural streams - minor streams (top width at flood stage < 100 ft)
1. Main Channels
a. clean, straight, full stage, no rifts or deep pools 0.025 0.030 0.033
b. same as above, but more stones and weeds 0.030 0.035 0.040
c. clean, winding, some pools and shoals 0.033 0.040 0.045
d. same as above, but some weeds and stones 0.035 0.045 0.050
e. same as above, lower stages, more ineffective slopes
and sections 0.040 0.048 0.055
f. same as "d" with more stones 0.045 0.050 0.060
g. sluggish reaches, weedy, deep pools 0.050 0.070 0.080
h. very weedy reaches, deep pools, or floodway
with heavy stand of timber and underbrush 0.075 0.100 0.150
2. Mountain streams, no vegetation in channel, banks usually steep, trees and brush along banks
submerged at high stages
a. bottom: gravels, cobbles, and few boulders 0.030 0.040 0.050
b. bottom: cobbles with large boulders 0.040 0.050 0.070
3. Floodplains
a. Pasture, no brush
1.short grass 0.025 0.030 0.035
2. high grass 0.030 0.035 0.050
b. Cultivated areas
1. no crop 0.020 0.030 0.040
2. mature row crops 0.025 0.035 0.045
3. mature field crops 0.030 0.040 0.050
c. Brush
1. scattered brush, heavy weeds 0.035 0.050 0.070
2. light brush and trees, in winter 0.035 0.050 0.060
3. light brush and trees, in summer 0.040 0.060 0.080
4. medium to dense brush, in winter 0.045 0.070 0.110
5. medium to dense brush, in summer 0.070 0.100 0.160
d. Trees
1. dense willows, summer, straight 0.110 0.150 0.200
2. cleared land with tree stumps, no sprouts 0.030 0.040 0.050
3. same as above, but with heavy growth of sprouts 0.050 0.060 0.080
4. heavy stand of timber, a few down trees, little
undergrowth, flood stage below branches 0.080 0.100 0.120
5. same as 4. with flood stage reaching branches 0.100 0.120 0.160
Source: http://www.fsl.orst.edu/geowater/
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 85
Appendix 9: Quality model syntax
(Source: Department of Water Quality Management, University of Wageningen)
/* River water quality model for the DO */ /* In case of river Tuul, Ulaanbaatar city, Mongolia */ /* MSc research */ /* WREM, ITC, 2008 */ WATER DO [8.0] mg/l ; Dissolved oxygen WATER BOD [5.0] mg/l ; Biochemical oxygen demand WATER NH4 [0.05] mg/l ; Ammonium PARM Beta [0.0001] (gO2/mg Chl)/ (W/m2) ; Oxygen production constant (photosynthesis) PARM fd [1.0] - ; Fraction of dissolved BOD PARM Kd [0.1] day ; BOD degradation rate constant PARM KDO [1.0] mgO2/l ; Monod constant DO inhibition BOD decay PARM KLmin [0.1] m/day ; Minimum oxygen mass-transfer rate in flowing system PARM KNDO [1.0] mgO2/l ; Monod constant DO inhibition nitrification PARM Knit [1.0] day ; Nitrification rate constant PARM OPTKL [1.0] - ; Option material transfer (0=stagnation 1=flowing) PARM TKd [1.05] - ; Temperature coefficient of BOD decay PARM TKL [1.024] - ; Temperature coefficient of oxygen mass-transfer PARM TKnit [1.05] - ; Temperature coefficient of nitrification PARM TSOD [1.06] - ; Temperature coefficient of sediment oxygen demand PARM Vs [1.0] m/day ; Settling velocity of BOD XT T [15.0] oC ; Water temperature XT SOD [2.0] g/m2, day ; Sediment oxygen demand XT SNH4 [0.0] g/m2, day ; Diffusion of NH4 concentration XT SBOD [0.0] g/m2, day ; Diffusion of BOD concentration XT I0 [0.00] W/m2 ; Solar radiation XT A [10.0] ug/l ; Algal biomass XT W [3.0] m/s ; Wind speed FLOW Q [10.0] m3/s ; Discharge FLOW As [60.0] m2 ; Cross-section of flow area FLOW Z [1.0] m ; Water depth U=ABS(Q/As); OS=14.652-0.41022*T+0.007991*T*T-0.000077774*T*T*T; IF (OPTKl==0)
{Kl20=0.0864*(8.43*W^0.5-3.67*W+0.43*W*W);} IF (W<1.82) {Kl20=0.37+0.09*W;} IF (OPTKl==1) {Kl20=2.33*U^0.67*Z^(-0.85);} IF (Kl20 <Klmin) {Kl20=Klmin;} KlT=Kl20*TKl^(T-20); KA=KlT/Z; BODU=BOD/(1-EXP(-Kd*5)); PDO=Beta*I0*A; SedDO=-SOD*TSOD^(T-20)/Z; REAR=KA*(OS-DO); Nitrif=-4.57*Knit*NH4*TKnit^(T-20)*DO/(DO+KNDO); BODox=-Kd*BODU*TKd^(T-20)*DO/(DO+KDO); k1(DO)=-KA; k0(DO)=KA*OS+PDO+SedDO+BODox+Nitrif; k1(BOD)=-Vs*(1-fd)/Z-Kd*TKd^(T-20)*DO/(DO+KDO); k0(BOD)=SBOD/Z; k1(NH4)=-Knit*TKnit^(T-20)*DO/(DO+KNDO); k0(NH4)=SNH4/Z;
A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA
DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 86
Appendix 10: A quality model setup
Upper CS – Zaisan Trapezoid Floor width=48.6 Slope=5.71 Max depth=1.5 Floor level=285.25 Surface level=292.6
Sonsgolon CS Trapezoid Floor width=44.5 Slope=5.65 Max depth=1.5 Floor level=265.44 Surface level=272.12
Songino (upper) CS Trapezoid Floor width=41.3 Slope=5.61 Max depth=1.5 Floor level=249.6 Surface level=255.73
Songino (down) CS Trapezoid Floor width=41.0 Slope=5.6 Max depth=1.5 Floor level=247.62 Surface level=253.69
Chicken farm CS Trapezoid Floor width=36.8 Slope=5.54 Max depth=1.5 Floor level=227.81 Surface level=233.2
Khadanhyasaa CS Trapezoid Floor width=33.5 Slope=5.5 Max depth=1.5 Floor level=211.97 Surface level=216.82
Altanbulag CS Trapezoid Floor width=26.6 Slope=5.4 Max depth=1.5 Floor level=178.3 Surface level=182
Lower CS – Extend Trapezoid Floor width=10 Slope=20 Max depth=1.5 Floor level=173.3 Surface level=177
BC upstream – Zaisan Q=0.033 BOD=6.0 DO=9.94 NH4-=0.7 Location: 643389 E, 5305700 N
BC - Sonsgolon Q=0.172 BOD=6.1 DO=8.48 NH4-=0.78 Location: 633139 E, 5303905 N
BC – Songino (upper) Q=0.155 BOD=6.0 DO=7.42 NH4-=1.51 Location: 626450 E, 5301560 N
BC – Songino (down) calibration Location: 625306 E, 5300735 N
BC – Chicken farm Q=2.025 BOD=26.9 DO=1.14 NH4-=1.94 Location: 619858 E, 5292251 N
BC – Khadanhyasaa Q=2.0 BOD=6.9 DO=4.4 NH4-=1.58 Location: 612478 E, 5289880 N
BC – Altanbulag Q=0.01 BOD=6.0 DO=7.5 NH4-=1.05 Location: 597134 E, 5283564 N
BC downstream – Artificial Q=-1.38 BOD=6.0 DO=7.5
NH4-=1.05 Location: 587593 E, 5301654 N
20 km far from Altanbulag
Section - Tuul 54 km long CS interpolated Min distance 10 m and max 200 m n=0.06
DP – CWTP Q=1.814 BOD=30.6 DO=4.41 NH4=29.33
DP – Bio Q=0.0172 BOD=30.6 DO=4.41 NH4=29.33
Calculation Starting
05.3.1-06.12.1 Time step=30’ Theta=0.55 Distance=200 Output = DO
SURFACE WATER QUALITY ASSESSMENT AND MODELLING
INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 87
Appendix 11: Required input data for the quality model
Type Abbreviation Source Typical value
DO CLEM
BOD5 CLEM Initial condition
NH4+ CLEM
DO CLEM
BOD5 CLEM Boundary condition
(system) NH4
+ CLEM
DO CLEM
BOD5 CLEM Boundary condition
(middle nodes) NH4
+ CLEM
KLmin Literature 0.1 m.day-1
TKL Literature 1.024
Kd Literature 0.1-0.3 day
Vs Literature < 1 m.day-1
fd Literature 0.8-1.0
KDO Literature 1 mg.l-1
TKd Literature 1.03-1.05
Knit Literature 0.1-1.0 day
TKnit Literature 1.05-1.10
KNDO Literature 1.0-2.0 mg.l-1
Beta Literature 0.0001-0.001
g day-1.mgChl-1.(W.m-2)-1
TSOD Literature 1.06
Parameters
OPTKL Literature 0=stagnant, 1=running
D Literature/estimate
T CLEM
SBOD Estimate
SNH4 Estimate
IO IWEC
A Estimate 10-200 ug.l-1
SOD Literature 0.5-2.0 g.m-2.day-1
External variable
W IWEC
Source: [Lijklema, et al., 1996]